11/j-hydroxyandrostenedione: a marker of adrenal function in hirsutism*t

Transcription

1 FERTILITY AND STERILITY Copyright" 1990 The American Fertility Society Printed on ocid-free poper in U.S.A. 11/j-hydroxyandrostenedione: a marker of adrenal function in hirsutism*t Robert W. Hudson, M.D., Ph.D.:j: Lynette J. Margesson, M.D.11 Heather A. Lochnan, M.D.+ II Barbara K. Strang, B.Sc.:j: F. William Danby, M.D.11 Susan M. Kimmett:j: Department of Medicine, Queen's University, and the Kingston General Hospital, Kingston, Ontario, Canada To assess the role ofthe adrenal glands in the development of hirsutism, levels of 11,8-hydroxyandrostenedione (11,8-0HA), 17a-hydroxyprogesterone (17-0HP), dehydroepiandrosterone sulphate (DHEAS), androstenedione ( 4 A), and free and total testosterone (T) were measured in 63 hirsute females and 30 control patients. Six of the hirsute patients had basal levels of 11,8-0HA and 17-0HP and responses to adrenocorticotropic hormone that were significantly greater than these values in controls and the other hirsute women. These women were designated as having an adrenal source for their hirsutism. Women with polycystic ovarian syndrome and idiopathic hirsutism had normal values of 11,8-0HA and 17-0HP. Levels of total and free T, DHEAS and 4 A were significantly higher than control values in all of the hirsute women. This study demonstrates that 11,8-0HA can be used as a marker to assess the adrenal contribution to hirsutism. Fertil Steril54:1065, 1990 The majority of hirsute females have evidence of increased androgen production, demonstrated by elevated serum levels of testosterone (T), T precursors, or androgen metabolites. 1 Often it is unclear whether the source of increased androgen production is the ovaries, the adrenal glands, or increased peripheral conversion of androgen precursors. 1 In particular, the role of the adrenal glands in the development of hirsutism often is unclear. 2 Dehydroepiandrosterone sulphate (DHEAS) has been Received March 21, 1990; revised and accepted August 13, * Supported by the Physicians of Ontario through a grant from the Physicians' Services Incorporated Foundation, Toronto, Ontario, Canada. t Presented in part at the 35th Annual Meeting of the Canadian Fertility and Andrology Society, Vancouver, British Columbia, Canada, November 9 to 11, i Division of Endocrinology. Reprint requests: Robert W. Hudson, M.D., Ph.D., Department of Medicine, Division of Endocrinology, Etherington Hall, Queen's University, Kingston, Ontario, Canada K7L 3N6. II Present address: Ottawa Civic Hospital, Ottawa, Ontario, Canada. 'IT Division of Dermatology. considered to be a marker of adrenal androgen production.3 However, the ovaries also can be a significant source of this steroid. 1 Therefore, a measure of DHEAS may not be discriminative in distinguishing those women in whom the adrenal glands are major contributors to their hirsutism. Similarly, androstenedione (a 4 A) is secreted by both the ovaries 4 and the adrenals. 1 Therefore, a measure of basal levels of this hormone does not distinguish those women with a significant adrenal contribution to their androgen excess. 11,8-Hydroxyandrostenedione (11,8-0HA) is the third most abundant steroid secreted by the adrenal glands, cortisol and DHEAS being more abundant.5 Its major precursor, a 4 A, is converted to 11,8-0HA by 11,8-hydroxylation.6-8 Because the 11,8-hydroxylase mediating this conversion is strictly an adrenal enzyme, 9 the ovaries do not secrete 11,8-0HA. A small quantity of this steroid, secreted by the adrenals, results from cortisol side-chain cleavage. 8 In some patients, a small proportion of circulating 11,8-0HA may derive from peripheral side chain cleavage of cortisol. 10 This normally amounts Hudson et al. The adrenal gland in hirsutism 1065

2 to <10% of the daily production of 11-0HA, however. Although the precise function of 11-0HA is not clear, it has been shown to have weak androgenic action. 11 Its secretion is under the control of adrenocorticotropic hormone (ACTH), its level rising after the administration of this peptide. Theoretically, then, this compound should represent 17 -oxosteroid production by the adrenal glands. This study was designed to see whether 11-0HA could be used as a marker of adrenal androgen production and to assess the role of the adrenal glands in the development of hirsutism. Patients Studied MATERIALS AND METHODS Sixty-three females, between 18 and 37 years of age, were referred for the assessment of hirsutism. Nine women were amenorrheic, whereas 16 continued to have regular periods. The rest had varying degrees of oligomenorrhea. Thirty women, 23 to 35 years of age, with no evidence of androgen excess or any condition known to affect the hypothalamic-pituitary-ovarian or adrenal axes, served as normal controls. Six women, 18 to 33 years of age, with no evidence of androgen excess and who were taking oral contraceptives (OCs), also were studied. Additionally, two women, ages 32 and 37, with newly diagnosed idiopathic primary adrenal insufficiency and one woman, 37 years old, with a newly diagnosed benign adrenal tumour causing Cushing's syndrome, were studied. Hormone Assays Serum levels of total T, total estradiol, luteinizing hormone (LH), follicle-stimulating hormone (FSH), and prolactin were measured by methods described previously. 12 Levels of cortisol, 17a-hydroxyprogesterone (17-0HP), DHEAS, and t:. 4 A were measured by specific radioimmunoassay kits. The intra-assay and interassay coefficientsofvariation (CVs) for all of the steroid and peptide hormone assays were <10%. 11-Hydroxyprogesterone was measured in duplicate in serum, using kits purchased from Intersci Diagnostics (Los Angeles, CA), according to the method of Polson et al. 13 The antiserum used was specific for 11-0HA. There was <1% cross-reac Hudson et al. The adrenal gland in hirsutism tion with t:. 4 A and <0.1% for all other steroids tested. The sensitivity of the assay for 11-0HA was 1.0 nm. The interassay and intra-assay CVs were<8%. Sex hormone-binding globulin (SHBG) was measured by a kit purchased from Nuclear Diagnostics (Ferndale, MI). The assay measures SHBG directly in serum utilizing an antihuman SHBG antibody raised in rabbits and a second, antirabbit globulin antibody. There is no detectable cross-reaction with human albumin, immunoglobulin G, thyroxine-binding globulin or transferrin. The sensitivity of the assay is 4 nm. The intra-assay and interassay CV s were <5% and 8%, respectively. Free T was determined by the method of Hammond et al. 14 Serum samples were incubated with 3H-T and 14 C-glucose and subjected to centrifugal ultrafiltration through a dialysis membrane (2,500 X g for 1 hour at 37.C). The percentages of free steroid were determined by comparing the ratio of 3H-T to 14 C-glucose in the ultrafiltrate with the corresponding ratio in the serum retained by the dialysis membrane. FreeT concentration was calculated from the percent free hormone and the total T concentration. Both intra-assay and interassay CVs were <10%. Adrenal Stimulation and Suppression Responses to ACTH were assessed by measuring serum levels of cortisol, T, t:. 4 A, DHEAS, 11-0HA, and 17-0HP at 30 minutes and just before the intramuscular (IM) administration of 0.25 mg of Synacthen (CIBA-Geigy Canada Ltd., Mississauga, Ontario, Canada), and at 15, 30, 45, 60, 90, and 120 minutes after the injection. The responses of these same steroids to glucocorticoid suppression were assessed by measuring their levels at 8:00 A.M. after the oral administration of 1.0 mg of dexamethasone (DEX) at 11:30 P.M. the previous night. Timing of Hormone Assays Serum for the hormone assays was obtained in the early follicular phase of those women having periods and at random in the amenorrheic women. Baseline levels were determined from samples obtained between 8:00 A.M. and 9:30 A.M. Determination of Radioactivity Radioactivity was determined using a Beckman scintillation analyzer (Beckman Instruments Inc., Fertility and Sterility

3 Table 1 11-Hydroxyandrostenedione and Cortisol Levels" Baseline 11-0HA Response to ACTH 11-0HA Pre Post Cortisol Pre Post After dexamethasone 11-0HA Cortisol Control 5.2 ± 1.9 (n = 30) 5.4 ± ± ± ± 198 (n= 6) 2.0 ± ± 15 (n=4) OCs Addison's Adrenal tumor 5.4± 1.5 < (n = 6) (n = 2) (n= 1) 5.8± 1.1 < ± 1.3 < ± ,335 ± (n = 3) (n = 2) 1.9± ± (n = 4) (n= 1) Oophorectomized women (n = 3): 11-0HA is 5.9 ± 0.7 nm. Mississauga, Ontario, Canada) as described previously.12 The significance of differences between mean values were determined using the Student's t-test. RESULTS Validation of Adrenal Source of HA The levels of 11,8-0HA are shown in Table 1. In the control patients, levels rose with ACTH and fell after DEX. Levels were undetectable in the patients with Addison's disease and did not respond to ACTH. The women on OCs and the oophorectomized women had levels of 11,8-0HA that were not different from normal. The patient with the adrenal tumor had markedly elevated levels that did not suppress with DEX. In each case, the dynamics of 11,8-0HA paralleled those of control. The Development of Subsets of Hirsute Women The hirsute women were divided into three subsets (Table 2). These subsets were developed on the basis of the validation of the adrenal source of 11,8-0HA, the patients' gonadotropin levels, and their ovarian ultrasound (US) results. (1) Adrenal Source Six women had 11,8-0HA levels significantly higher than control values. These patients had normal LH/FSH ratios and normal appearing ovaries on US. They were designated as having an adrenal cause for their hirsutism. (2) Polycystic Ovarian Syndrome Twenty-three women were designated as having polycystic ovarian syndrome (PCOS) because of early follicular phase ratios of LH to FSH > 2.5:1 and US evidence of enlarged ovaries with multiple cysts (Table 2). These patients all had 11,8-0HA levels that were not different from normal. (3) Idiopathic Hirsutism Thirty-four women had normal LH:FSH ratios, no US evidence of ovarian cysts, and normal levels of 11,8-0HA. These women were designated as having idiopathic hirsutism because they could not be identified as having a clear ovarian or adrenal source for their hirsutism. Table 2 Control and Hirsute Patients Studied Age Body mass index LH FSH Ovarian US" 11-0HA y kgfm' IU/L nm Control (n = 30) All hirsute (n = 63) Adrenal (n = 6) PCOS (n = 23) Idiopathic hirsutism (n = 34) 29 ± ± ± 7 25 ± ± ± ± 7.9b 26 ± ± ± 9b 5.3± ±3.6 N 5.2 ± ± 7.3b 8.2 ± ± ± ±3.5 N 10.0 ± 3.5b 21.3 ± 9.4b 7.1 ± 4.8 c 6.0 ± ± ±3.2 N 5.7 ± 1.8 Ultrasound appearance of ovaries: N, normal; C, multiple cysts. b Significantly different from control value; P < Hudson et al. The adrenal gland in hirsutism 1067

4 Table 3 Hormone Levels in Controls and Hirsute Patients Total T FreeT SHBG DHEAS il 4 A 17-0HP Control (n = 30) 1.1 ± ±0.01 All hirsute (n = 63) 2.6 ± 0.7" 0.11 ± 0.06" Adrenal (n = 6) 3.4 ± 0.19" 0.16 ± 0.05" PCOS (n = 23) 2.3 ± 1.2" 0.10 ± 0.05" Idiopathic hirsutism (n = 34) 2.6 ± 0.45" 0.10±0.07" Significantly different from control values; P < nm I'm 65± ± ± ± ± 17" 10.0 ±5.0" 4.0 ± 1.7" 2.4 ± ± 18" 12.6 ± ± 2.1" 3.6 ± 0.39" 20± 19" 8.9 ±4.8" 4.2 ± 1.8" 2.0 ± ± 18" 10.2 ±4.7" 3.4 ± 1.5b 2.4 ± 1.2 b Significantly different from control values; P = nm Hormone Levels The levels of total T, free T, DHEAS, and!!. 4 A were significantly greater than normal in all of the hirsute women, whereas SHBG levels were lower than control values (Table 3). The six women with higher than normal levels of 11,8-0HA also had levels of 17-0HP that were greater than the levels in both the control patients and in PCOS and idiopathic hirsutism. Response to ACTH The responses to ACTH are shown in Figure 1. The responses of 11,8-0HA and 17-0HP were significantly greater than normal only in the adrenal group. The responses in the adrenal group also were significantly greater than the responses of the 15 "'<l 'C 1:.. 10 I 5 ) ) t f1 f;a l OL tj 12 f1.s 8 a. J: t I pa ( 4-0 CONTROL PCO ADRENAL IDIOPATHIC before before after befol8 after before after n=5 n=4 n=s n= ,... SOOT..: :J: 12 'J: 8 Figure 1 The peak responses of cortisol, 111'1-0HA, il 4 A, DHEAS, and 17-0HP to ACTH (0.25 mg Synacthen IM) in control women and in women with hirsutism. PCOS women and the women in the idiopathic hirsutism group who underwent this test. The cortisol response in the adrenal group was the lowest of all the groups but overlapped with normal (Fig. 1). Although basal levels of DHEAS were greater than normal in the hirsute women, there was not a consistent rise in DHEAS during the 2 hours of the ACTH test in any of the groups studied. The!J. 4 A response to ACTH was greatest in the adrenal group but not different, significantly, from the response of the other two hirsute groups. There was not a consistent rise in total or free T in any of the groups studied. Prolactin Serum PRL levels were 7.4 ± 3.6, 6.2 ± 1.3, 10.8 ± 5.7, and 9.4 ± 5.5 ug/l for control, adrenal, PCOS, and idiopathic hirsutism, respectively. None of the values for the hirsute women was different, significantly, from the controls. Effect of Therapy Adrenal Group Three of the women in the adrenal group were given 0.25 mg of DEX orally, at bedtime, as their only treatment. They had a substantial fall in 11,8-0HA, cortisol, and 17 -OHP after 3 to 5 months of therapy (Table 4). Total and free-t levels also were lower than pretreatment values. Although there was a fall in DHEAS and!!. 4 A, these levels remained detectable. Sex hormone-binding globulin values did not change significantly with DEX alone. The other three women in the adrenal group were treated with a combination of DEX and an OC, Demulen 30 (G. D. Searle and Co. of Canada, Ltd., Oakville, Ontario, Canada) or Ortho 777 (Ortho Pharmaceuticals Ltd., Don Mills, Ontario, 1068 Hudson et al. The adrenal gland in hirsutism Fertility and Sterility

5 Table 4 The Effect of Therapy on Hormone and SHBG Levels Cortisol llfj-oha Total T Adrenal (n = 3) Pre 415 ± ± ± 0.6 Post 84± ± ±0.5 (Pvalue) (0.001) (0.004) (0.01) PCOS (n = 12) Pre 492 ± ± ± 1.2 Post 502 ± ± ± 0.5 (Pvalue) (NS)" (NS)" (0.001) Idiopathic hirsutism (n = ll)b Pre 516 ± ± ± 1.7 Post 482 ± ± ± 1.0 NS, not significant. Canada) because of their need for contraception. These women had significant falls in all steroid levels and elevations in SHBG. Polycystic Ovarian Syndrome Twelve women with PCOS were treated with OCs only (Demulen 30 or Ortho 777). After 3 to 5 months of treatment, they had significant falls in total and free T, DHEAS, and tl. 4 A. These falls were associated with a significant increase in SHBG levels. There were no changes in cortisol or 11-0HA levels. Although 17-0HP values were lower than pretreatment levels, they were not different from levels in the control patients. The other women in this group were either treated with a combination of OCs and spironolactone or chose not to be treated. Idiopathic Hirsutism Eleven women in idiopathic hirsutism, who had had tubal ligations, were treated with spironolactone, alone, in a dose of 100 mg/d. There was no significant change in levels of any of the hormones or in SHBG, as assessed after 3 to 5 months of therapy (Table 4). The other women in this group either were treated with a combination of OCs and spironolactone or chose not to be treated. During the time of the study, there was not a significant change in body mass index in any of the groups of women (data not shown). DISCUSSION The results of the current study have confirmed that 11-0HA is strictly an adrenal secretory prod- nm FreeT SHBG DHEAS 17-0HP 0.12 ± ± ± ± ± ± ± ± ± ± 0.7 (0.005) (NS) (0.02) (0.05) (0.004) 0.11 ± ± ± ± ± ± ± ± ± ± 0.1 (<0.001) (<0.001) (0.02) (0.01) (0.02) 0.10 ± ± ± ± ± ± ± ± ± ± 1.8 b P value not significant. uct. The levels measured in the control patients are similar to levels reported previously by other investigators In the normal individuals, levels of 11-0HA rose in response to ACTH and fell nearly to undetectable levels after the administration of DEX. Its levels were low and failed to respond to ACTH in the patients with primary adrenal insufficiency. Levels were virtually identical to that of control patients, in the oophorectomized women, and those using OCs. The qualitative responses of 11-0HA to ACTH and DEX were identical to those of cortisol. As a result, it would appear that this steroid can be used as a marker of adrenal 17-ketosteroid production. Of the 63 hirsute women studied, only 6 had basal levels of 11-0HA that were greater than normal. These women also had greater than normalll-oha responses to ACTH. Moreover, they had basal levels of 17-0HP and responses to ACTH that were greater than normal. These elevated basal levels of 11-0HA and 17 -OHP andresponses to ACTH were significantly greater than these values in the other 57 hirsute females. It is likely, therefore, that these 6 women have a different etiology for their hirsutism than do the other hirsute women. Their hormone responses are consistent with them having an adrenal cause for androgen excess. Although basal levels of DHEAS and tl. 4 A also are greater than normal in the adrenal group, they were not significantly different from these values in the other hirsute women. Moreover, the responses ofdheas and tl. 4 A to ACTH were not significantly different in this group of women from the other hirsute females. nm Hudson et al. The adrenal gland in hirsutism 1069

6 Twenty-three women had early follicular phase gonadotropin levels and US findings consistent with a diagnosis of PCOS. Of note is the fact that the steroid levels of these women were indistinguishable from the adrenal group if only T, DHEAS, or f:.. 4 A levels were considered. Moreover, the responses of DHEAS and f:.. 4 A to ACTH in PCOS did not distinguish them from the adrenal group. These findings reconfirm that a measure of these steroids does not distinguish, uniformly, those women hirsute due to ovarian causes from those secondary to adrenal causes. Similarly, the 34 women with idiopathic hirsutism could not be distinguished from the adrenal group except by their normal levels of ll{j-oha and 17-0HP. It is acknowledged that ACTH tests were not carried out in all of the women in the PCOS and idiopathic hirsutism groups. It is possible that some women in each of these groups may have had an adrenal contribution to their hirsutism despite normal basal levels of ll{j-oha. These women may have had an excessive ll{j-oha and/or 17-0HP response to ACTH as has been described for patients with nonclassical congenital adrenal hyperplasia.16 One might be alerted to this possibility in these women if their total and free-t levels did not suppress to control levels with OCs. An ACTH test should then be performed. Further studies in this area are needed to explore this possibility. The data from this study have confirmed that the majority of hirsute females have elevated levels of total T. There is some overlap, however, with the range described as normal by many laboratories.1 13 Mean free-t levels, however, did not overlap with normal in the patients investigated in this study. On the basis of this study, it appears that a measurement of free T, by centrifugal ultrafiltration, is an accurate method of demonstrating increased androgen production in hirsute women. The lower than normal SHBG levels in all of the hirsute females were consistent with their elevated free T concentrations. A measure of SHBG appears to be an accurate assessment of androgen action. The uniformly low levels of this protein in the hirsute patients is consistent with the known effect of androgens on SHBGY The levels rose, as expected, with OCs. The fall in free T in the patients in PCOS treated with OCs is consistent with this rise in SHBG. Interestingly, there was no rise in SHBG, despite a fall in free T, in the patients in the adrenal group, treated with DEX for up to 3 to 5 months. The ex- planation for this lack of a rise is not clear. It may be related to the short duration of treatment, the small number of patients or, perhaps, to a suppressive effect ofdex on SHBG.18 Also of interest is the overall lack of a change in any of the hormones or in SHBG in the patients treated with spironolactone. Other investigators have reported that spironolactone may be associa ted with a fall in total T but no change in non SHBG bound Tor SHBG Although some of the patients in the current study had a fall in total and free T, others had a rise, whereas in the rest, no difference was noted. When the data were taken as a group, there was no change in any of the parameters measured when spironolactone was used alone. This finding is consistent with its primary action being through competitive inhibition of androgen binding to the receptor in the hair follicle It may also inhibit 17a-hydroxylase and 17,20 lyase activities, competitively inhibit the binding of T to SHBG, and increase the metabolism oft.22 This study has demonstrated that ll{j-oha can be used as a marker of adrenal17-ketosteroid production and may be useful in identifying those women with hirsutism who have a major adrenal contribution. At least in this study, about 10% of hirsute women have a significant adrenal source of excessive androgen production. It would be logical to treat these women by suppressing their adrenal glands. Further work is needed to determine whether the women with PCOS or idiopathic hirsutism can be shown, by stimulation with ACTH or by attempting to suppress their adrenal contribution with DEX to have subtle, increased adrenal androgen production. Studies to this end currently are in progress. Acknowledgment. The authors thank Mrs. Carolyn Ferguson for her help in performing the study and Ms. Debbie Rawlins for secretarial assistance in preparing the manuscript. REFERENCES 1. Moltz L, Schwartz V: Gonadal and adrenal androgen secretion in hirsute females. In Clinics in Endocrinology and Metabolism, Androgen Metabolism in Hirsute and Normal Females, Vol. 15, Edited by R Horton, RA Lobo. London, W.B. Saunders Company, 1986, p Kirschner MA, Jacobs JB: Combined ovarian and adrenal vein catheterization to determine the sites of androgen overproduction in hirsute women. J Clin Endocrinol Metab 33:199, Hudson et al. The adrenal gland in hirsutism Fertility and Sterility

7 3. Buvat J, Siame-Mourot C, Fourlinnie JC, Lemaire A, Buvat-Herbaut M, Hermand E: Androgens and prolactin levels in hirsute women with either polycystic ovaries or "borderline ovaries." Fertil Steril38:695, Ryan KJ, Petro Z: Steroid biosynthesis by human ovarian granulosa and thecal cells. J Clin Endocrinol Metab 26:46, Goldzieher JW, Beering SC: Metabolism of lltl-hydroxyandrostenedione, adrenosterone and hydrocortisone to urinary 11-oxy-17-ketosteroids. J Clin Endocrinol Metab 29: 171, Cohn GL, Mulrow PJ: Androgen release and synthesis in vitro by human adult adrenal glands. J Clin Invest 42:64, Adadevah BK, Engel LL, Shaw D, Gray CH: Metabolism ofprogesterone-4-14 C by adrenal tissue from a patient with Cushing's syndrome. J Clin Endocrinol Metab 25:784, Hudson RW, Killinger DW: The in vitro biosynthesis of lltl-hydroxyandrostenedione by human adrenal homogenates. J Clin Endocrinol Metab 34:215, Gower DB: Properties and subcellular location of enzymes involved in steroidogenesis (and the role of cytochrome P450). In Biochemistry of Steroid Hormones, Edited by HLJ Makin. Oxford, Blackwell Scientific Publications, 1975, p Gower DB: Catabolism and excretion of steroids. In Biochemistry of Steroid Hormones, Edited by HLJ Makin. Oxford, Blackwell Scientific Publications, 1975, p Howard E: A complementary action of corticosteroid and dehydroepiandrosterone on the mouse adrenal with observations in the sensitivity of reproductive tract structures to dehydroepiandrosterone and 11-hydroxyandrostenedione. Endocrinology 65:785, Hudson RW, Perez-Marrero RA, Crawford VA, McKay DE: Hormonal parameters in incidental varicoceles and those causing infertility. Fertil Steril45:692, Polson DW, Reed MJ, Franks S, Scanlon MJ, James VHT: Serum lltl-hydroxyandrostenedione as an indicator of the source of excess androgen production in women with polycystic ovaries. J Clin Endocrinol Metab 66:946, Hammond GL, Nisker JA, Jones LA, Siiteri P: Estimate of the percentage of free steroid in undiluted serum by centrifugal ultrafiltration dialysis. J Biol Chern 255:5023, Goldzieher JW, Pena A, Aivaliotis MM: RIA of androstenedione, testosterone and lltl-hydroxyandrostenedione after chromatography on LIPIDEX J Steroid Biochem 9: 169, New MI: Polycystic ovarian disease and congenital and late-onset adrenal hyperplasia. In Endocrinology and Metabolism Clinics of Normal America. Polycystic Ovarian Disease, Vol. 17, Edited by DK Mahajan. Philadelphia, W.B. Saunders Company, 1988, p Easterling WE, Talbert LM: Serum testosterone levels in the polycystic ovary syndrome. Am J Obstet Gynecol 12: 385, Stanczyk FZ, Petra PH, Jenner JW, Novy MJ: Effect of dexamethasone treatment on sex steroid-binding protein, corticosteroid-binding globulin, and steroid hormones in cycling rhesus macaques. Am J Obstet Gynecol 151:464, Lobo AR, Shoupe D, Serafini P, Brinton D, Horton R: The effect of two doses of spironolactone on serum androgens and anagen hair in hirsute women. Fertil Steril43:200, Givens JR: Treatment of hirsutism with spironolactone. Fertil Steril43:841, Loriaux DL, Menard R, Taylor A, Pita JC, Santen R: Spironolactone and endocrine dysfunction. Ann Intern Med 85:630, Caminos-Torres R, MaL, Snyder PJ: Gynecomastia and semen abnormalities induced by spironolactone in normal men. J Clin Endocrinol Metab 45:255, 1977 Hudson et al. The adrenal gland in hirsutism 1071