Stimulation of gonadal steroid synthesis bychronic excess of adrenocorticotropin inpatients with adrenocortical insufficiency.

H Vierhapper, … , P Nowotny, W Waldhäusl

J Clin Invest. 1986;77(4):1063-1070. https://doi.org/10.1172/JCI112405.

Analysis of 24-h urinary steroid excretion was performed by capillary gas chromatography insix patients (five men, one woman) with adrenocortical insufficiency. Ten healthy subjects(five men, five women) served as controls. A complete absence of all 21-hydroxylatedsteroid metabolites was seen in patients with adrenocortical insufficiency, whereas theexcretion of several steroids lacking hydroxylation in the 21-position (pregnenolone,pregnenetriol, and 11-ketoandrosterone) was markedly increased. In addition, the presenceof 11 beta-hydroxyandrosterone was confirmed by mass-spectrometry in the urine of threepatients. This pattern of steroid excretion was unchanged in patients with adrenocorticalinsufficiency, both after stimulation by 1-24 adrenocorticotropin (ACTH) and after short-term(3-d) suppression with dexamethasone. We conclude that patients with adrenocorticalinsufficiency present a pattern of steroid excretion characterized by the absence of 21-hydroxylated metabolites. In the absence of functional adrenocortical tissue, long-termpathologically elevated concentrations of ACTH apparently stimulate early steps of steroidsynthesis, most likely in the gonads. In addition, the presence of 11-hydroxylated steroidmetabolites (11-ketoandrosterone, 11 beta-hydroxyandrosterone) in the urine of patientswith adrenocortical insufficiency demonstrates that chronic ACTH excess in this disordermay induce some activity of 11 beta-hydroxylase, an enzyme not found in the gonads underphysiological conditions.

Research Article

Find the latest version:

http://jci.me/112405/pdf

Pdf

Stimulation of Gonadal Steroid Synthesis by Chronic Excess ofAdrenocorticotropin in Patients with Adrenocortical InsufficiencyH. Vierhapper, P. Nowotny, and W. WaldhauslDivision of Clinical Endocrinology and Diabetes Mellitus, I. Medizinische Universitdtsklinik, A-1090 Wien, Austria

Abstract

Analysis of 24-h urinary steroid excretion was performed bycapillary gas chromatography in six patients (five men, onewoman) with adrenocortical insufficiency. Ten healthy subjects(five men, five women) served as controls. A complete absenceof all 21-hydroxylated steroid metabolites was seen in patientswith adrenocortical insufficiency, whereas the excretion of severalsteroids lacking hydroxylation in the 21-position (pregnenolone,pregnenetriol, and 11-ketoandrosterone) was markedly increased.In addition, the presence of 1 1j-hydroxyandrosterone was con-firmed by mass-spectrometry in the urine of three patients. Thispattern of steroid excretion was unchanged in patients with ad-renocortical insufficiency, both after stimulation by 1-24 adren-ocorticotropin (ACTH) and after short-term (3-d) suppressionwith dexamethasone. Weconclude that patients with adreno-cortical insufficiency present a pattern of steroid excretion char-acterized by the absence of 21-hydroxylated metabolites. In theabsence of functional adrenocortical tissue, long-term patholog-ically elevated concentrations of ACTH apparently stimulateearly steps of steroid synthesis, most likely in the gonads. Inaddition, the presence of 11-hydroxylated steroid metabolites(1 1-ketoandrosterone, 1 1 fl-hydroxyandrosterone) in the urine ofpatients with adrenocortical insufficiency demonstrates thatchronic AC1THexcess in this disorder may induce some activityof 11,6-hydroxylase, an enzyme not found in the gonads underphysiological conditions.

Introduction

The simultaneous evaluation of major urinary steroid metab-olites by capillary gas chromatography (1, 2) comprehensivelycharacterizes steroid production in man. Since a differentiationof metabolites derived from adrenocortical or gonadal steroidsynthesis is not possible from determinations performed underbasal conditions, dynamic tests of adrenocortical function (i.e.,stimulation by adrenocorticotropin [ACTH] or suppression byexogenous glucocorticoids) are used to characterize the contri-bution of the adrenal cortex to overall steroid secretion. Theinterpretation of the results thus obtained relies on the assump-tion that these maneuvers interfere with adrenal, but not withgonadal, steroid synthesis. Although it is undisputed that ACTH

Please address correspondence to Dr. Vierhapper, Division of ClinicalEndocrinology and Diabetes Mellitus, I. Medizinische Universittitsklinik,Lazarettgasse 14, A-1090 Wien, Austria.

Receivedfor publication 20 August 1984 and in revisedform 26 No-vember 1985.

J. Clin. Invest.© The American Society for Clinical Investigation, Inc.0021-9738/86/04/1063/08 $ 1.00Volume 77, April 1986, 1063-1070

predominantly influences the function of the adrenal cortex, anadditional effect on gonadal steroid production has not yet beenexcluded. Patients with adrenocortical insufficiency provide anexcellent model from which to study the effect of markedly in-creased endogenous concentrations of ACTHon gonadal steroidproduction in the absence of functioning adrenocortical tissue.In the present investigation we have compared steroid excretionrates in patients having adrenocortical insufficiency with thoseof healthy controls, both under basal conditions as well as afteradditional stimulation with exogenous ACTH 1-24 and aftersuppression with dexamethasone.

Methods

Healthy subjects. The subjects were five healthy female volunteers, aged19-27 yr, who had not taken oral contraceptives for at least 6 mo, andfive healthy male volunteers, aged 21-28 yr. All subjects were within± 10%of their ideal body weight (according to Metropolitan Life InsuranceTables). The aim and potential risks of the investigation were explainedto each subject and written consent was obtained in each case. In healthywomen, the study was begun between the fifth and the seventh day ofthe menstrual cycle. Following the collection of a 24-h urine for thedetermination of steroid excretion, an indwelling catheter was insertedinto an antecubital vein and 0.25 mg synthetic ACTH(1-24 ACTH,Synacthen, 0.25 mg = 100 IU; CIBA-GEIGY Ltd. Basle, Switzerland)was administered at 8:00 a.m. as an intravenous bolus. This was followedby a continuous infusion (3 ml/h; t = 8 h) containing 0.25 mg 1-24ACTHin 24 ml of 0.9% NaCI solution. On days 3-6 of the experimentthe subjects were treated with oral dexamethasone (0.5 mg four times aday; Schering A.G., Berlin, Federal Republic of Germany), but not withintravenous ACTH. A 24-h urine (8:00 a.m.-8:00 p.m.) was collectedfor the determination of urinary steroid excretion on day 1 (basal), day2 (ACTH), and day 5 (dexamethasone) of the experiment. Blood samplesfor the determination of plasma concentrations of cortisol, 21 -deoxy-cortisol and 11 -ketoandrostenedione, pregnenolone, and 17-hydroxy-pregnenolone were obtained before the infusion of ACTH. Plasma con-centrations of cortisol were also determined after the administration ofACTHand of dexamethasone.

Patients with adrenocortical insufficiency. Six patients with adreno-cortical insufficiency (five men, aged 26-48 yr, and one postmenopausalwoman, aged 55 yr) consented to participate in the study. '3'I-cholesterolscanning failed in all patients to reveal the presence of functional adre-nocortical tissue. Adrenocortical insufficiency was of autoimmune originin four patients and due to bilateral adrenalectomy for Cushing's diseasein two cases. Pituitary surgery had not been performed on any of thesepatients. The patients had been on continuous glucocorticoid (37.5-50.0mg cortisone acetate/d divided into two doses) and mineralocorticoid(0.05-0.1 mg 9a-fluoro-hydrocortisone/d) replacement therapy for 2-17 yr before this investigation. Nevertheless, supranormal concentrationsof ACTH(> 100-> 1,000 pg/ml; normal <60 pg/ml) were found in eachpatient when estimated during routine controls on an outpatient basis.For the purpose of this study the patients were admitted to the hospitaland taken off their individual substitution therapy. Consecutively, elec-trolyte equilibrium was monitored daily and maintained by intravenousadministration of 0.9% saline. 84 h after the last dose of the respectivesubstitution therapy the patients received an intravenous bolus and an

Steroid Excretion in Adrenocortical Insufficiency 1063

infusion of 1-24 ACTHas described above for the healthy volunteers.Subsequently each patient was given 0.5 mgdexamethasone four timesa day and 0.05 mg9a-fluoro-hydrocortisone/d by mouth for 3 d. A 24-h urine (08:00-08:00 h) for the determination of steroid excretion wascollected under basal conditions (i.e., from 60 to 84 h after withdrawalof substitution therapy) during ACTHstimulation and on the third dayof reinstituted glucocorticoid (dexamethasone) therapy. Blood samplesfor the determination of plasma concentrations of cortisol were drawnat the beginning of each urine-collecting period and at the end of theACTHinfusion. Concentrations of ACTH, of 2 1-deoxycortisol, of 11-ketoandrostenedione, of pregnenolone, and of 1 7-hydroxypregnenolonewere obtained before the' infusion of ACTH. Concentrations of ACTHwere also estimated after the administration of dexamethasone.

Reference steroids. Reference steroids were purchased from SteraloidsInc., Wilton, NH; and from Makor Chemicals, Jerusalem, Israel. Puritywas checked by controlling melting point in each case.

Radioimmunoassays. For the simultaneous radioimmunologic de-termination of cortisol, 2 1-desoxycortisol, and 1 1-ketoandrostenedione,2.0 ml of plasma were diluted with 8.0 ml H20. To determine recoveries20,000 dpm of [3H]cortisol (NET 396, specific activity 115 Ci/mmol;NewEngland Nuclear, Boston, MA), 1,000 dpmof 21-[3H]desoxycortisol(specific activity 43 Ci/mmol; Amersham International, Amersham,Buckinghamshire, England), and 10,000 dpm of [3H]DOC (specific ac-tivity 42 Ci/mmol; Amersham International) were added to each sample.[3HJDOC was used to estimate the recovery of 11-ketoandrostenedione,since the latter is not available in a labeled form and shows a chromato-graphic behavior in the chromatographic system we employed which isidentical to that of DOC. Subsequently the samples were extracted with150 ml of cold CH2Cl2 and the organic extract was washed once with10 ml of cold 0.1 NNaOHand twice with 10 ml of cold H20. Separationof the steroids was achieved by thin-layer chromatography (benzene/acetone = 75:25). For the simultaneous determihation of pregnenoloneand 17-hydroxypregnenolone, 2 ml of plasma were diluted with 8 ml ofH20. To determine recoveries 10,000 dpmof [3H]pregnenolone (specificactivity 12 Ci/mmol; Amersham Corp.) and 17-[3H~hydroxypreg-nenolone (specific activity 19'Ci/mmol; Amersham Corp.) were addedto each sample. Subsequently the samples were extracted with 150 mlof cold CH2Cl2 and the organic extract was washed with 10 ml of coldH20. Separation of steroids was achieved by two-dimensional thin-layerchromatography (I: benzene/acetone, 75:25; II: ethylacetate/cyclohexane,55:45). Radioimmunoassays were performed in triplicate using antiseraagainst cortisol and 2 1-desoxycortisol obtained from Dr. Vecsei, Divisionof Pharmacology, University of Heildelberg, Heidelberg Federal Republicof Germany (3). Antisera for the determination of pregnenolone and 17-hydroxypregnenolone were obtained from Steranti Research Ltd., St.Albans, England; and from Radioassay Systems Laboratories, Inc., Car-son, CA, respectively. Since no specific antiserum was available for thedetermination of 1 l-ketoandrostenedione, this steroid was determinedusing the cross-reactivity (12.5%) of an antibody raised against andro-stenedione (Nr. XXA001; Steranti Research Ltd.). Cross-reactivity ofthis antibody against DOCis <0.1%. Furthermore, the behavior of an-drostenedione and 1 1-ketoandrostenedione permits complete separationin the used chromatographic system. Urinary concentrations of cortisoland plasma concentrations of ACTHwere determined as previously re-ported (4).

Urine processing. Urine samples (5 ml) of 24-h collections were acid-ified to pH 5.2 by addition of 0.2 N HCI. ,B-glucuronidase/arylsulfatase(20,000 Fisher units, Calbiochem-Behring Corp., La Jolla, CA) was added,the hydrolysis taking place at 37°C. Hydrolysis did not affect pH values.After a total hydrolysis period of 24 h, 50,000 dpm [3Hjcortisol were

added to estimate recovery. The unconjugated steroids were extractedwith three times 20 ml of ethylacetate and the combined extract waswashed once with 20 ml of 0.1 N NaOHand at least twice with 20 mlof double distilled water until neutrality was reached. Anhydrous sodiumsulfate (5-10 g) was then added to each sample. After 12 h the solventwas distilled off at 350C at reduced pressure. The residue was transferredwith 3 x 1 ml ethylacetate into a centrifuging tube and the solvent wasevaporated in a stream of dry nitrogen at 350C.

Derivatization (5, 6). Samples were dissolved in 100 ,d of a solution

of methoxyamine hydrochloride (2% in pyridine; Pierce Chemical Co.,Rockford, IL) and heated at 80'C for 60 min. After evaporation withdry N2 at 300C, 100 Ml of a mixture of trimethylsilylimidazole, NO-bis(trimethylsilyl)acetamide, and trimethylchlorosilane (3:3:2, TRI-SILTBT; Pierce Chemical Co.) was added and the mixture was heated to60C for 20 h. After evaporation to dryness by dry N2 (250C, overnight),2 ml of dichloromethane and 1.5 ml of 0.1 NH2SO4were added to eachsample, the mixture was shaken, and the aqueous (lighter) phase waspipetted off. The organic layer was washed once with I ml of doubledistilled water. If the pH of the washing was >7, the organic layer wasdiluted with 1 ml 0.1 N H2SO4. The organic layer was then diluted with1 ml CH2Cl2 and washed with H20 until the washings were neutral.Sodium sulfate was added to the organic solutions and the samples wereshaken for 20 min and centrifuged. The clear solutions were transferredinto reactivials; the solvent was evaporated with dry N2 at room tem-peratur: and the residue was dissolved in 100 ,l of a mixture of ethyl-acetate/n-hexane (9:1) containing 100 ng/Al of alkanes n-C24H50 andn-C32H66

Recovery. 20 Ml of each sample were measured in a fl-counter (PackardInstruments Co., Delft, Netherlands). Mean recovery (n = 75) was55.7+5.5%.

Gas chromatography. Gas chromatography was carried out using aPackard Instrument Co. 428 gaschromatograph equipped with a 50 mOV-101 wall-coated open tubular column (0.25 mmi.d., film thickness0.25 Mm, pretested at 180.000 effective plates), a solid injection device,and a flame ionization detector. N2 was used as a carrier gas with a flowrate of -0.8 ml/min. After sample injection (300°C), the temperatureof the oven was immediately programmed between 180°C and 265°Cat 0.40C/min followed by an isothermal run for 15 min at end temper-ature.

Steroid identification. Steroids were identified by comparison of re-tention times expressed as methylene units as determined by intrapolationbetween those of the alkanes n-C24H50 and n-C32H66; (Table I). Trivialnames of steroid metabolites determined by gas chromatography and ofthe other steroids mentioned in this report are given in Table I.

Mass spectrometry. Positive identification of I 1-hydroxyandrosteronewas attempted by gas chromatography-mass spectrometry using a 5995gas chromatography-mass spectrometry system equipped with a 25-mfused silica capillary column coated with OV101 stationary phase (Hew-lett-Packard Co., Palo Alto, CA). The GC injector was maintained at300°C and the oven was programmed from 180°C to 260°C at a rateof I °C/min. Temperatures of the analyzer, transfer line, and ion sourcewere kept at 1800C, 280°C, and 150°C, respectively, and the electronenergy was 70 eV. To increase sensitivity it was necessary to determine

I 1-hydroxyandrosterone by selected ion monitoring, the instrumenthaving been set to monitor the ions at m/e 479 (M+), m/e 449 (M+- 30), m/e 448 (M+- 31), and m/e 358 (M+- 121).

Calculation of steroid excretion. An automatic integrator (3358 A;Hewlett-Packard Co.) was used for the calculation of the area under therespective peaks. Response factors (f) for the two n-alkanes (f C24 andf C32) and for each reference steroid standard (f St) were determined bytheir injection in varying amounts. These response factors were checkedand adjusted regularly (i.e., after every 15th injection of biological sam-ples) by derivatizing and injecting defined mixtures of the reference com-pounds. Response factors for the two n-alkanes were practically identical(f C24, 1.05±0.18; f C32, 1.07±0.20) over the complete series of injec-tions. It was unnecessary to introduce additional correction factors forthe increasing elution temperatures. The excreted amounts of individualsteroids were calculated by the following formula:

mg/24 h = area X f C24 X f St X 100/recovery(%) X dilution,

where area equals measured peak area; f C24 equals response factor forhydrocarbon C24 (provides for the actually injected amount); f St equalsresponse factor for individual steroid; recovery equals the yield of thecomplete treatment of the samples before gas chromatography by mea-suring recovery of [3H]cortisol; and dilution equals the aliquotization oforiginal urinary samples.

Sensitivity and precision. The sensitivity of the procedure varied ac-cording to the gas chromatography conditions used. Applying standard

1064 H. Vierhapper, P. Nowotny, and WWaldhdusl

Table I. Trivial Names and Gas-chromatographic Characteristics of Evaluated Standard Steriod Compounds*

Methylation unitsNo. Steroid Trivial name (as MO-TMSderivatives)

Sa-Androstan-3a-ol- 17-one5,B-Androstan-3a-ol- 17-one5-Androsten-3,B-ol- 17-oneSa-Androstan-3,B-ol- 17-one5-Androsten-313,173-diol5a-Androstan-3a-ol-1 1, 17-dione1,3,5(10)-Estratrien-3-ol- 17-one5,B-Androstan-3a-ol- 11, 17-dione4-Androsten-3,17-dione1,3,5( 10)-Estratrien-3,173-diol4-Androsten- 173-ol-3-one5,B-Pregnan-3a-ol-20-one5a-Androstan-3a- 1 3-diol- 17-one5,B-Androstan-3a- If3-diol- 17-one50-Pregnan-3,3-ol-20-one5a-Pregnan-3a,20a-diol5,B-Pregnan-3a,20a-diol5-Pregnen-30-ol-20-one5(3-Pregnan-3a, 17a,20a-triol5-Androsten-3,B-16a, 17a-triol4-Pregnen-3,20-dione5-Androsten-30,16a, 17(-triol503-Pregnan-3a, 17a,2 1-triol-20-one5,B-Pregnan-3a,2 1-diol-20-one1,3,5( 10)-Estratrien-3, 16a, 1 7/3-triol5a-Pregnan-3a,2 1-diol-20-one5a-Pregnan-3a, 17a,2 I-triol-20-one53-Pregnan-3a, 17a,20a-triol- 1l-one5-Pregnen-3fl, 17a,20a-triol5,B-Pregnan-3a, 17a,2 I-triol-I 1 ,20-dione5(3-Pregnan-3a,2 1-diol-1 1,20-dione5(3-Pregnan-3a, 1 1(,2 1 -triol-20-one5a-Pregnan-3a, 1 1,,2 1-triol-20-oneSa-Pregnan-3a, 17a,2 1-triol-I 11,20-dione50-Pregnan-3a, 1 1(, 17a,2 I-tetrol-20-oneSa-Pregnan-3a, 1 1(, 17a,2 I-tetrol-20-one5,B-Pregnan-3a, 17a,20a,2 1-tetrol- 1l-one5,B-Pregnan-3a, 1 7a,20,B,2 I-tetrol- 1 -one5,B-Pregnan-3a, 11(, 17a,20,B,21 -pentol5-Cholesten-3,B-ol5/3-Pregnan-3a, 113, 17a,20a,2 1-pentol

n-C24-AlkaneAndrosteroneEtiocholanoloneDehydroepiandrosteroneEpiandrosteroneAndrostenediol1 l-KetoandrosteroneEstrone1 1-KetoetiocholanoloneAndrostenedioneEstradiolTestosteroneEpipregnanolone1 IB-Hydroxyandrosterone1 lB-HydroxyetiocholanolonePregnanoloneAllo-PregnanediolPregnanediolPregnenolonePregnanetriol17a-AndrostenetriolProgesterone17,B-AndrostenetriolTH-"-S")'TH-"DOC"EstriolAllo-TH-"DOC"Allo-TH-"S"PregnanetriolonePregnenetriolTH-"E"TH-"A"

oTH-"B"Allo-TH-"B"Allo-TH-"E"TH-(4'F"Allo-TH"F"a-Cortolone,B-Cortolone(3-CortolCholesterola-Cortoln-C32-Alkane

* Trivial names of other steroids mentioned in this paper:4-Androsten- 1,B-ol-3,17-dione4-Androsten-3,11,17-trione5-Pregnen-3j, 17a-diol,20-one4-Pregnen- 1I3, 17a-diol-3,20-dione4-Pregnen- I7a-ol-3,20-dione4-Pregnen- 17a,2 I-diol-3,11,20-trione4-Pregnen- 113, 17a,2 I-triol-3,20-dione4-Pregnen-1 1(3,2 1-diol-3,20-dione4-Pregnen,2 1-ol-3,20-dione4-Pregnen- 17a,2 I-diol-3,20-dione

1 1,B-Hydroxyandrostenedione1 l-Ketoandrostenedione17a-Hydroxypregnenolone2 1-Desoxycortisol17a-Hydroxyprogesterone (" 170H-P")Cortisone ("E")Cortisol ("F")Corticosterone ("B")Desoxycorticosterone ("DOC")1 1-Desoxycortisol ("S")

23456789

10111213141516171819202122232425262728293031323334353637383940414243

24.0025.0925.2825.6825.8325.9226.0126.0126.1426.19/26.2426.3826.4726.8726.9627.1327.1527.4227.6227.6727.9528.1628.20/28.2528.4128.6728.7628.7628.9129.0229.0229.4729.6629.7530.0030.1230.2530.2730.3730.4930.7730.7730.8931.2232.00

Steroid Excretion in Adrenocortical Insufficiency 1065

conditions sensitivity was -1 ng per injected sample corresponding toabout 0.02 mgof steroid per 24 h. The coefficient of variation for multipledeterminations (n = 8) of the whole spectrum of standard compoundswas 12±3% (range, 10-20%).

Statistics. Data are presented as mean±SD. Student's t test (two-tailed) for matched and for unmatched pairs was used for statistical eval-uation.

Results

In unsubstituted patients with adrenocortical insufficiency,plasma concentrations of cortisol were 1. 1±0.5 ug/dl (controls,6.8±2.2 ug/dl). Concentrations of 1 1-ketoandrostenedione andof 21-deoxycortisol were 64±18 ng/dl (controls, 279±139 ng/dl) and 2.0±0.3 ng/dl (controls, 5.1 ± 1.0 ng/dl) respectively, andthe concentrations of pregnenolone and 17-hydroxypregneno-lone were 38.2±14.2 ng/dl (controls, 53.3±6.2 ng/dl) and12.9±6.4 ng/dl (controls, 45.0±18.9 ng/dl), respectively. Thus,when compared with healthy controls, mean estimated concen-trations of cortisol, 1 I-ketoandrostenedione, 21 -deoxycortisol,pregnenolone, and 17-hydroxypregnenolone in patients withadrenocortical insufficiency were 13, 23, 39, 72, and 29%, re-spectively. In patients with adrenocortical insufficiency, plasmaconcentrations of cortisol were unchanged by the administrationof ACTH (1.5±0.7 Atg/dl; controls, 24.0±10.6 gg/dl) and ofdexamethasone (0.6±0.2 ,ug/dl; controls, 0.8±0.2 ,g/dl). Plasmaconcentrations of ACTHin patients with adrenocortical insuf-ficiency were 968±410 pg/ml (range, 462-1670 pg/ml) afterwithdrawal of substitution therapy and 151 ± 144 pg/ml (range,54-429 pg/ml) after 3 d of dexamethasone by mouth.

Steroid excretion rates of healthy men, healthy women, andof patients with adrenocortical insufficiency are given in TableII. Comparable excretion rates of cholesterol were found in pa-tients with adrenocortical insufficiency and in healthy controls.Urine samples of patients were characterized by a lack of all 21-hydroxylated metabolites (TH-"A", TH-"B", allo-TH-"B", TH-"DOC", allo-TH-"DOC", TH-"S", allo-TH-"S", TH-"E", allo-TH-"E", TH-"F", allo-TH-"F", a-cortol, f3-cortol, a-cortolone,and fl-cortolone, Fig. 1). In addition, patients with adrenocorticalinsufficiency presented with a reduced excretion of androsterone,etiocholanolone, dehydroepiandrosterone, androstenediol, 11 (,-hydroxyetiocholanolone, 1 3l-hydroxyandrosterone, 11 -keto-etiocholanolone, pregnanediol, and allopregnanediol, whereasthe mean excretion of pregnanetriol (3.92±3.07 mg/24 h), preg-nenetriol (2.22±2.47 mg/24 h), pregnenolone (1.85±2.47 mg/24 h), and 1 1-ketoandrosterone (1.34±1.64 mg/24 h) was in-creased as compared with healthy controls. In regard to the lowexcretion rates, both in patients with adrenocortical insufficiencyand in healthy controls, a comparison of the excreted amountswas not possible for the remaining compounds listed in TablesI and II. Finally, the excretion rates of cortisol in patients withadrenocortical insufficiency were <0.5 ,ug/24 h after withdrawalof substitution therapy, during stimulation by ACTH, and after3 d of dexamethasone by mouth. Whereas the administrationof ACTHand of dexamethasone induced the expected changesin the excretion of steroid metabolites in healthy men andwomen, the respective excretion rates remained unchanged inpatients with adrenocortical insufficiency (Table II).

Gas-chromatographic analysis indicated the presence of 11-hydroxyandrosterone in the urine of three patients with adre-nocortical insufficiency. This was confirmed by additional mass

spectrometry using selected ion monitoring. As shown in Fig. 2for one single case the chosen characteristic ions were readilyidentified, both in a mixture of derivatized reference compoundsand in the patient's urine.

Discussion

The prerequisite for the simultaneous estimation of a largenumber of urinary steroid metabolites has been provided by thedevelopment of glass capillary gas chromatography (1, 2). Thevariety of techniques and hydrolytic procedures used by differentlaboratories implies that the only valid reference material is amaterial obtained using that very method. In order to study thesteroid excretion rates in patients with adrenocortical insuffi-ciency, it was therefore pivotal to establish reference data in agroup of healthy subjects. These data are included in this report.

Although the relative share of the gonads and the adrenalcortex in overall steroid production can easily be studied in pa-tients with adrenocortical insufficiency, few such data have beenreported as yet in this disorder (7-9).

Under physiological conditions the enzymes 2 1-hydroxylase,1 lB-hydroxylase, and 1 8-hydroxylase are unique to the adrenalcortex, whereas other enzymes, such as 3,B-steroid dehydrogenase,17a-hydroxylase, and the desmolases are shared by the adrenalsand by the gonads and probably are controlled by a single gene(10). The lack of 2 1-hydroxylated steroid metabolites observedin the urine of our patients can therefore be easily deduced fromthe differences in the enzymatic equipment of the adrenal cortexas compared with that of the gonads, whereas the presence oftwo 1 1-hydroxylated metabolites, 1 I11B-hydroxyandrosterone and1 1-ketoandrosterone, and the elevated excretion of pregneno-lone, pregnanolone, and pregnenetriol requires further discus-sion.

Although adrenal scanning is believed to be superior to othernoninvasive techniques in the detection of adrenal rest tissue(11), the negative outcome of '311-cholesterol scanning per semay not completely rule out this possibility in our patients. Ad-ditional evidence against a residual adrenal function is, however,provided by the lack of urinary cortisol and by plasma cortisolconcentrations close to the lower limit of the used radioim-munological method. Finally, the increased excretion of somemetabolites not requiring 2 1-hydroxylation (e.g., pregnanetrioland pregnenetriol) in the absence of all urinary 2 1-hydroxylatedmetabolites and a similarly distorted relationship among the es-timated plasma steroids implies that the site of steroid hormoneproduction in our patients is characterized by a pattern of ste-roidogenesis different from that of normal adrenal tissue.

In the absence of functional adrenal tissue the presence of1 1-hydroxylated steroid metabolites must be due to an extra-adrenal 1 1-hydroxylase. Such a speculation may be based onthe previous demonstration of an extraadrenal 21 -hydroxylase(12) and of a hepatic 313-steroid dehydrogenase (10). However,in regard to the increased excretion of early steroid metabolitesmentioned above, a gonadal origin of 1 l3-hydroxylated steroidsin our patients appears more likely. The minor alterations ingonadotropin secretion seen in patients with adrenocortical in-sufficiency (13) are not likely to account for these pronouncedchanges in steroid excretion.

Under the appropriate experimental conditions, ACTH, thetropic hormone of the adrenal, may stimulate both rat and hu-man testicular cells at the molecular level in terms of RNAsyn-

1066 H. Vierhapper, P. Nowotny, and W. Waldhdusl

(71 00%)-o 0-e

O - o -

° °V °

01 en a-Un- 0

+l +1 +1

00 0

+ + 1- +1 I-_ _ 0ocN- 6

-N ^ O -

II11 °I100°1

^ eNI0% ~NOFNNO 'C_N ,,,NN,0 -0O o ooo 0- -o.eO .0O. .0o.66O 6OO6OO66O66O66OoooO

-II--+ + I -I- -I- I+ I0%0oo_-oO N66 6 6

° 1

0-oS Inol

InOoi"roo0%N£ t0 0t

1 +1

n en

4F- i0 N e N t

++1 +1 +l-N-00 %

- oo0- 0

V- en0 %

1. 1.'Os

InI 660

N0. Os 0

00%6ClC5C

+1 +1 +1- 0 -

C

t

4+ 4+0%l v F en en %o

i N -00- 0

+1+1+1 +1in % 0 %0eiIR Ci Cien Cj0 6

N _4

+1N

6

0 0

+100Cli6

N_ -_ Ct = 0 C - M 0 W0o V

066666 coma-oo+1 +1 +1 +1 +1

-00 n 0, enI 66

- - -

I 1 6 -; 6

00 e

II I 1+1o~~~~~~~~~~~~~~l en

00̂ ^ t

(1 "o - O eM - oo 't 4t oo - O -_Q -0 N_ q-00 -: O 00-

+1+1 +1+1 + +1Un o 0 00 N- e-.00 --: -:en6O -0 -- 6

4++ co') +4 4+O e t m N 0- 0% 0 Cn

- o 't - d-t -R- - en - _(6 6 6 66 6_-6 6 666

+1 +1 +1 +1 + +1rCN b o en N

140O --: "t14en m 0 0

N Oen O6 6

008

00.6

kn 00-ef> O~

I1°-

en en r-t" o _ t }4 bumr- oOeq - O nO

+1 +1 +1 +1 +1 +1_0 N-_Nt o- NFl ~ -

0o- O o0Ov0Fo0 -O tO tO~~t O O ~ O^ O ~ O0 cq

+ +1 +1 +1 +1 +1 +10% 0 N e mr

I -~,6-S 6

N'IT

II 011 111iOe -000o-0

1+ 1+11

66tof 66ooo£_ o - o -I-I O

.6

e

N +II 01111

00

I0

7

zII +1

' t I- t eRT e

+

00

0

00

6i

'tO C 000Orm% D 0 %0£_ 00 n 00 c 00o0 O en6 - 6 - o o - - - o o

+ +I +l +I +I +I +1 +1 +1_ 0 ID in tno o -

; o a ao0 r- 00

O -C i e-; C i - o

comen C- so en C1 _ T

-Cf0 0en 0U 0 0e

+1 + +I +I +I +I +I00 It e rNO- 0

6 -io _r . en

\0 N0 0

InI

6611I

00

0

-0ON4 0% 00

+l +1 +1n n

r- e -

'tO %0R---0 0 0 en"C- Oq e - -C ei " -

+ +l +I +I +I +I +1 +1 +1C1 -; "S eM o oo t -;C 00-r e t In

- O a -

4+O IC 0 e t 00 O

C) ~o0 -t - - --! .- . . .

0 (O a C> 0 a 0 0+++ I +1 +1 +1ON N enR e -0 - O O -

0- 00- Cl4 -

+1 +1O N-6 6

O 0

C)0~~ C C~ ~ ~ C r C

00.0 00t t

L~0V )00C

- y ~H~~~~oo-=..)-

2C

Exg8

00 N-00

. :.

00 000 06 6v v

L

00eq

r-06

. ei0 0

00 %0

0 0

+ +~0 NI

0 000q O~

00 -Cl

-0

_ _

66v v

.0-

.& 8

C)C)

.0s

ot .

Le

C)

. D~

_ .2C) 0

0 .0

to

._

O cn

2>s0.-O=

_) -

C.O

. a0 .^

30 0

0 0>

0. COez

Q .3L Q*

Cq.

C..

*-0

Cl

00

elI II 01I

O

+ +C O

M 00oo- 0

+

6

Steroid Excretion in Adrenocortical Insufficiency 1067

CHOLESTEROL (41) PREGNENETRIOL(30)

PREGNENOLONE(19) '-l 177-HYDROXY- - DHEA (4) ANDRO;TENEDIOL (6)I PREGNENOLONE

PREGNANOLONE(16)4' PROGESTERONE (22) * 17o(-HYDROXY- - R ANDROTENEDIONE(10) - TESTOSTERONE (12)PROGESTERONE

[21 - HYDROXYLASE]

CORTICOSTERONE 21-DESOXYCORTISOL..

I CDESOXYCORTICOSTERRONE 21-DESOX1L\XPREGNANEDIOL(18) TH-A' (32) ..-. 'I\

TH-'B' (33), allo-TH-'B'(34) .TH-'S' (24) 1113-HYDROXY-ANDRO-TH-'DOC' (25) allo-TH-'S' (28) STENEDIONEo~lo-TH-'DOC' (27) .I

PREGNANETRIOL(20)

CORTISONE CORTISOL

TH-'E' (31) allo-TH-'E' (35) TH-'F' (36)d.-Cortolone (38) B-Corto)one (39) oL -Cortol (42)

e 4I1-KETO 11-KETO

ETIOCHOLANOLONEANDROSTERONE(9) (7)

t ~~t

411 B-HYDROXY

ANDROSTcRONE(14)t

allo-TH-'F' (37)B-Cortol (40)

I 1 -HYDROXYETIOCHOLANOLONE

(15)t

1-KETO-ANDRO-STENEDIONE

ETIOCHOLANOLONEANDROSTERONE(3) (2)

Figure 1. Simplified scheme of steroidogenesis in patients with adrenocortical insufficiency. The numbers in brackets refer to the numbering sys-tem of steroid standards given in Table I.

thesis and of steroid hormone synthesis ( 14). It appears possible In the adrenal cortex the presence of elevated concentrationsthat concentrations of ACTHelevated, as in our patients, for of ACTHfor prolonged periods of time has been shown to alterperiods of several years may influence gonadal steroidogenesis the relative activities of the various steroid hydroxylases (15).in adrenocortical insufficiency. Chronic overexposure to ACTHapparently modifies gonadal

ST.SUM OF MONITORED ION ABUNDANCESFull Scale 500

A. ~ ~~~~~~~~~~

ION 479.30

M~~~~~~~~~~~~~~ .

F.S.w 5

ION *449.30F.S.= 13M+ -30

ION 448.30 M-3F.S.m 34

M+3

ION *358.20 M-2F.. 28

091%~~~~~~~~~~~~~1

37 38 39 40 41 42 43 44 45 46 47 48

MINUTES

P.SUM OF MONITORED ION ABUNDANCESFull Scale - t00

.1: :

; %%

ION 479.30F.S. ' I

ION ' 449.30F.S.= M+- 30

ION - 448.30 M+3F.S.= 1 -31

ION z 358.20F.S. * 2 M+-121f\,

36 37 38 39 40 41 42 43 44 45 46 47 48

MINUTES

Figure 2. Identification of I 1-hydroxyandrosterone by gas chromatography-mass spectrometry (selected ion monitoring) in a mix-ture of derivatized reference compounds (St.) and in the urine of a patient with adrenocortical insufficiency (P.). Plots of the fol-lowing four ions are shown: m/e 479 (M+), m/e 449 (M+- 30), m/e 448 (M+- 31), and m/e 358 (M+ - 129). The abscissarepresents the retention time (minutes).

1068 H. Vierhapper, P. Nowotny, and W. Waldhdusl

steroidogenesis in an analogous manner, thus explaining the in-creased excretion of early steroid metabolites not requiring 21-hydroxylation. This would also imply that the attempt to qualifythe origin of any given steroid as gonadal or adrenal by its absenceor presence in patients with adrenocortical insufficiency (9) isan oversimplification. Steroid excretion rates remained un-changed in our unsubstituted patients during the administrationof 1-24 ACTHand of dexamethasone, indicating that the ad-ditional administration of ACTH or the short-term partialsuppression of endogenous ACTHby 2 mgdexamethasone for3 d is apparently of little relevance in this pathological situation.Concerning the presence of 1 1-hydroxylated metabolites, i.e., of11 3-hydroxyandrosterone and 11 -ketoandrosterone, in the urineof patients with adrenocortical insufficiency, it should be re-membered that some testicular tumors in patients with congen-ital adrenal hyperplasia and in adrenalectomized patients withCushing's syndrome possess both 1 13-hydroxylase and 2 1-hy-droxylase activity in vitro ( 16-19). The question of whether thesetumors arise from regular testicular tissue (20) or from adrenalremnants within the testes (21) is as yet unsettled (19), but pro-longed exposure to markedly elevated concentrations of ACTHhas generally been implied in their pathogenesis (17). Thus, dur-ing prolonged overexposure to ACTH, the human gonad is ca-pable of producing 1 1-hydroxylated steroids.

Since the side chain at C17 found in C21 steroids orients thereduction of the double bond at C4/C5 in the Sf direction (22),C1903 steroids with the 5:l-configuration (i.e., 1 1-ketoetiochol-anolone, and 1 Il3-hydroxyetiocholanolone) are metabolites ofC21 precursors such as cortisol. Compounds with the 5a config-uration (i.e., 11 -ketoandrosterone, 1 (3-hydroxyandrosterone) arederived from 1 l/3-hydroxyandrostenedione and from 1 I-ke-toandrostenedione (22-24) via side-chain cleavage of 2 1-deoxy-cortisol. The synthesis of these steroids, therefore, requires theactivity of 1 l3-hydroxylase, but not of 2 1-hydroxylase. The as-sumed presence of 11 3-hydroxylase in the gonads of our patientsthus not only explains the observed excretion of C 1 -oxygenatedsteroids, but also the preponderance of the 5a- vs. the 5(-con-figurated metabolites.

The estimated plasma concentrations of 21 -desoxycortisoland 1 -ketoandrosterone are also in keeping with the presenceof 1 l-hydroxylase activity in our patients, and it is of note thatthe reduction in the plasma concentrations of both steroids wasless pronounced than that of cortisol. In regard to the elevatedexcretion rates of 1 -ketoandrosterone, pregnanetriol, pregne-netriol, and pregnenolone, elevated plasma concentrations ofthe respective precursor steroids (i.e., 1 -ketoandrostenedione,21-desoxycortisol, pregnenolone, and 17-hydroxypregnenolone)might have been expected. This was clearly not the case andthere is no ready explanation for this observation, although onemight point out that plasma concentrations of steroids do notnecessarily reflect the excretion rates of their metabolites, sinceboth values also depend on the metabolic clearance rate of therespective precursor (25).

Under physiological conditions both the adrenal cortex andthe gonads contribute to the urinary excretion of 11 -desoxy- 17-ketonic steroid metabolites (etiocholanolone, androsterone), al-though direct glandular secretion of both compounds is small(26). In our patients with adrenocortical insufficiency the lackof adrenal precursors apparently accounts for the diminishedexcretion of etiocholanolone and androsterone. The latter is inkeeping with the 65% reduction in plasma androsterone as re-

ported by others (26).

In conclusion, patients with adrenocortical insufficiencypresent a pattern of steroid excretion characterized by the com-plete absence of 2 1-hydroxylated, but not of 1 1-hydroxylated,metabolites. The obtained data suggest that chronic ACTHex-posure may result in the activation of 1 1f-hydroxylase in thehuman gonad.

Acknowledgment

The valuable technical assistance of Ms. E. Nowotny is gratefully ac-knowledged.

References

1. Luyten, J. A., and G. A. F. M. Rutten. 1974. Analysis of steroidsby high resolution gas-liquid chromatography. II. Application of urinarysamples. J. Chromatogr. 91:393-406.

2. Pfaffenberger, C. D., and E. C. Horning. 1975. High resolutionbiochemical gas-chromatography. Determination of human steroid me-tabolites using glass open tubular capillary columns. J. Chromatogr. 1 12:581-594.

3. Milewicz, A., P. Vecsei, S. Korth-Schutz, D. Haack, A. Rosler, K.Lichtwald, S. Lewicka, and G. von Mittelstaedt. 1984. Development ofplasma 2 1-deoxycortisol radioimmunoassay and application to the di-agnosis of patients with 2 1-hydroxylase deficiency. J. Steroid Biochem.21:185-191.

4. Waldhausl, W., K. Herkner, P. Nowotny, and P. Bratusch-Marrain.1978. Combined 17a- and 18-hydroxylase deficiency associated withcomplete male pseudohermaphroditism and hypoaldosteronism. J. Clin.Endocrinol. Metab. 46:236-246.

5. Chambaz, E. M., and E. C. Horning. 1969. Conversion of steroidsto trimethylsilyl derivatives for gas phase analytical studies: reaction ofsilylating reagents. Anal. Biochem. 30:7-24.

6. Shackleton, C. H. L., and J. W. Honour. 1976. Simultaneousestimation of urinary steroids by semiautomated gas chromatography.Investigation of neo-natal infants and children with abnormal steroidsynthesis. Clin. Chim. Acta. 69:267-283.

7. Abraham, G. E., and Z. H. Chakmakjian. 1973. Serum steroidlevels during the menstrual cycle in a bilaterally adrenalectomized woman.J. Clin. Endocrinol. Metab. 37:581-587.

8. Worgul, T. J., R. J. Santen, W. Samojlik, and S. A. Wells. 1982.How effective is surgical adrenalectomy in lowering steroid hormoneconcentrations? J. Clin. Endocrinol. Metab. 54:22-26.

9. Urban, M. D., P. A. Lee, J. P. Gutai, and C. J. Migeon. 1980.Androgens in pubertal males with Addison's disease. J. Clin. Endocrinol.Metab. 51:925-929.

10. Bongiovanni, A. M. 1981. Acquired adrenal hyperplasia: withspecial reference to 3,B-hydroxysteroid dehydrogenase. Fertil. Steril. 35:599-607.

11. Beierwaltes, W. H., J. C. Sisson, and B. Shapiro. 1984. Diagnosisof adrenal tumors with radionuclide imaging. Spec. Top. Endocrinol.Metab. 6:1-54.

12. Winkel, C. A., M. L. Casey, R. J. Worley, J. D. Madden, andP. C. MacDonald. 1983. Extraadrenal steroid 21 -hydroxylase activity ina womanwith congenital adrenal hyperplasia due to steroid 2 1-hydrox-ylase deficiency. J. Clin. Endocrinol. Metab. 56:104-107.

13. Vierhapper, H., W. Waldhausl, and P. Nowotny. 1982. Gonad-otrophin-secretion in adrenocortical insufficiency: impact of glucocor-ticoid substitution. Acta Endocrinol. 101:580-585.

14. Villee, D. B., and J. Fisher. 1975. The influence of ACTHonsteroid synthesis in normal and abnormal testicular tissues. Adv. EnzymeRegul. 13:317-328.

15. Ney, R. L., R. N. Dexter, W. W. Davis, and L. D. Garren. 1967.A study of the mechanism by which adrenocorticotroic hormone main-tains adrenal steroidogenic responsiveness. J. Clin. Invest. 46:1916-1924.

16. Dominguez, 0. V. 1961. Biosynthesis of steroids by testiculartumors complicating congenital adrenocortical hyperplasia. J. Clin. En-docrinol. Metab. 21:663-674.

Steroid Excretion in Adrenocortical Insufficiency 1069

17. Hamwi, G. J., G. Gwinup, J. H. Mostow, and P. K. Besch. 1963.Activation of testicular adrenal rest tissue by prolonged excessive ACTHproduction. J. Clin. Endocrinol. Metab. 23:861-869.

18. Fore, W. W., T. Bledsoe, D. M. Weber, R. Akers, and R. T.Brooks. 1972. Cortisol production by testicular tumors in adrenogenitalsyndrome. Arch. Intern. Med. 130:59-63.

19. Radfar, N., F. C. Bartter, R. Easley, J. Kolins, N. Javadpour,and R. J. Sherins. 1977. Evidence for endogenous LH suppression in aman with bilateral testicular tumors and congenital adrenal hyperplasia.J. Clin. Endocrinol. Metab. 45:1194-1204.

20. Landing, B. H., and E. Gold. 1951. The occurrence and signifi-cance of Leydig cell proliferation in familial adrenal cortical hyperplasia.J. Clin. Endocrinol. Metab. 11:1436-1453.

21. Wilkins, L., W. Fleischmann, and J. E. Howard. 1940. Macro-genitosomia precox associated with hyperplasia of the androgenic tissueof the adrenal and death from corticoadrenal insufficiency. Endocrinology.26:385-395.

22. Keutmann, E. H., and W. B. Mason. 1967. Individual urinary17-ketosteroids of healthy persons determined by gas chromatography:biochemical and clinical considerations. J. Clin. Endocrinol. Metab. 27:406-420.

23. Bradlow, H. L., and T. F. Gallagher. 1959. Metabolism of 11-hydroxy-A4-androstene-3,17-dione in congenital adrenal hyperplasia. J.Clin. Endocrinol. Metab. 19:1575-1580.

24. Goldzieher, J. W., and S. C. Beering. 1969. Metabolism of 1 Ij3-hydroxyandrostenedione, adrenosterone and hydrocortisone to urinary1 1-oxy 17-ketosteroids. J. Clin. Endocrinol. Metab. 29:171-178.

25. Samojlik, E., M. A. Kirschner, D. Silber, G. Schneider, andN. H. Ertel. 1984. Elevated production and metabolic clearance rates ofandrogens in morbidly obese women. J. Clin. Endocrinol. Metab. 59:949-953.

26. Meikle, A. W., J. D. Stringham, D. E. Wilson, and L. I. Dolman.1979. Plasma Sa-reduced androgens in menand hirsute women: role ofadrenals and gonads. J. Clin. Endocrinol. Metab. 48:969-975.

1070 H. Vierhapper, P. Nowotny, and W. Waldhausl