Adrenal dynamic responses to physiologicand pharmacologic adrenocorticotropichormone stimulation before and afterovarian steroid modulation in women withpolycystic ovary syndrome

Frank Gonzalez, M.D.,* Lillie Chang, M.D.,† Theresa Horab, R.N.,*Frank Z. Stanczyk, Ph.D.,† Kent Crickard, M.D.,* and Rogerio A. Lobo, M.D.†‡

School of Medicine and Biomedical Sciences, State University of New York at Buffalo, Buffalo, New York,and University of Southern California School of Medicine, Los Angeles, California

Objective: To test the hypothesis that in women with polycystic ovary syndrome (PCOS), adrenal cyto-chrome P450c 17a activity is different after physiologic vs. pharmacologic ACTH stimulation and thatovarian activity promotes adrenal hyperactivity that is different after physiologic vs. pharmacologic ACTHstimulation.

Design: Prospective controlled pilot study.

Setting: Reproductive endocrinology unit of an academic medical center.

Patient(s): Six women with PCOS who had adrenal hyperandrogenism were compared with four women withnormal ovulation.

Intervention(s): Adrenal dynamic blood sampling was performed before and after 6 months of GnRH agonistadministration.

Main Outcome Measure(s): Comparison of physiologic and pharmacologic ACTH-stimulated levels ofprogesterone, 17-hydroxyprogesterone, and androgens before and after ovarian steroid modulation.

Result(s): In women with PCOS, exaggerated responses of androstenedione and 11b-hydroxyandrostenedi-one as well as elevated ratios of 17-hydroxyprogesterone to progesterone and of androstenedione to 17-hydroxyprogesterone after physiologic ACTH stimulation did not persist after GnRH-agonist administration.Three of the six women with PCOS had an increased response of androstenedione and a ratio of andro-stenedione to 17-hydroxyprogesterone that were.2 SD above the mean of those in the women with normalovulation after pharmacologic ACTH stimulation; this finding persisted after GnRH-agonist administration.

Conclusion(s): In women with PCOS, increases in adrenal androgen sensitivity after physiologic ACTHstimulation reflected in both arms of cytochrome P450c 17a activity may be influenced by ovarian activity.However, 17,20-lyase hyperactivity in a subset after pharmacologic ACTH stimulation may be an intrinsicadrenal disorder. (Fertil Sterilt 1999;71:439–44. ©1999 by American Society for Reproductive Medicine.)

Key Words: Polycystic ovary syndrome, adrenal dynamic testing, GnRH agonist, adrenal androgen sensi-tivity, 17,20-lyase hyperactivity

Adrenal androgen excess in women withpolycystic ovary syndrome (PCOS) has beenattributed to adrenal androgen hyperrespon-siveness to ACTH or to an intrinsic adrenalenzyme dysfunction (1). This adrenal androgenhyperresponsiveness has been observed afterpharmacologic ACTH administration or phys-iologic ACTH release in response to cortico-tropin-releasing hormone administration (2–4).The latter phenomenon is an indicator of in-

creased adrenal androgen sensitivity (1, 5) andalso is more reflective of the in vivo circum-stance because circulating levels of ACTH arenormal in women with PCOS (6).

Adrenal enzyme dysfunction characterizedby a deficiency is a relatively rare phenomenonand is an entity that should be differentiatedfrom PCOS (7–9). In contrast, enhanced adre-nal enzymatic activity has been proposed tooccur in association with PCOS.

Received June 17, 1998;revised and acceptedOctober 8, 1998.Supported in part by aResearch CompetitionAward from the Universityat Buffalo ResearchFoundation, Buffalo, NewYork.Presented at the 42ndAnnual Meeting of TheSociety for GynecologicInvestigation, Chicago,Illinois, March 15–18, 1995.Reprint requests:Frank Gonzalez, M.D.,Department of Gynecologyand Obstetrics, Children’sHospital of Buffalo, 219Bryant Street, Buffalo, NewYork 14222 (FAX: 716-888-3833).* Department ofGynecology andObstetrics, School ofMedicine and BiomedicalSciences, State Universityof New York at Buffalo.† Department of Obstetricsand Gynecology, Universityof Southern CaliforniaSchool of Medicine.‡ Present address:Department of Obstetricsand Gynecology, Collegeof Physicians andSurgeons of ColumbiaUniversity, New York, NewYork.

FERTILITY AND STERILITY tVOL. 71, NO. 3, MARCH 1999Copyright ©1999 American Society for Reproductive MedicinePublished by Elsevier Science Inc.Printed on acid-free paper in U.S.A.

0015-0282/99/$20.00PII S0015-0282(98)00469-5

439

Previous investigators (10) have reported a dysregulationof the ovarian cytochrome P450c 17a-hydroxylase–17,20-lyase enzyme complex (cytochrome P450c 17a) character-ized by an increase in the activity of 17a-hydroxylase rela-tive to that of 17,20-lyase. Data supporting this abnormality(10) revealed a markedly increased 17-hydroxyprogesterone(17-OHP) response to ovarian dynamic testing. Dysregula-tion of the adrenal counterpart of this enzyme complex hasbeen suggested as the cause of excess adrenal androgensecretion in women with PCOS (11).

In a study by Azziz et al. (12), increases in the activity ofeither arm of adrenal cytochrome P450c 17a could not beconfirmed in most of the hyperandrogenic women evaluated.In contrast, we have published data supporting the conten-tion that some women with PCOS may have increasedadrenal activity of only the 17,20-lyase arm of this enzymecomplex (9). Contributing to this controversy is the fact thatonly pharmacologic doses of ACTH were used to evaluatethis enzyme activity. Moreover, the use of physiologic dosesof ACTH for this purpose to simulate the normal circulatinglevels of ACTH observed in vivo in women with PCOS (6)has never been attempted.

There are data to suggest that ovarian activity also influ-ences adrenal androgen excess in women with PCOS. Sig-nificant reductions in the basal levels of serum DHEAS andthe post-ACTH responses of other androgens have beenobserved after GnRH agonist (GnRH-a) administration inwomen with PCOS who have elevated adrenal androgenlevels but not in those who have normal adrenal androgenlevels (2, 13–15). In addition, we have reported enhance-ment of adrenal 17,20-lyase activity by the administration ofestrogen during GnRH-a treatment of women with PCOSwho have adrenal androgen excess (16).

Thus, we designed a pilot study to test a two-part hypoth-esis. First, we proposed that adrenal cytochrome P450c 17aactivity in women with PCOS is different when assessed byphysiologic compared with pharmacologic ACTH stimula-tion. Second, we explored the hypothesis that ovarian activ-ity may promote adrenal hyperactivity in women with PCOSthat may be different in response to physiologic vs. pharma-cologic ACTH stimulation.

MATERIALS AND METHODS

SubjectsSix women with PCOS were selected for the study. They

all had elevated basal levels of serum DHEAS as determinedwith a well-established RIA (17, 18) with a normal range forwomen of reproductive age between 70mg/dL and 300mg/dL. They all exhibited the classic features of oligomen-orrhea and hirsutism and had no evidence of virilization orpelvic masses. The mean (6SEM) number of menses was4.76 0.8 per year and the mean (6SEM) Ferriman-Gallweyhirsutism score was 10.56 1.2. All the women had with-

drawal bleeding after progesterone administration. All thewomen had normal levels of prolactin, TSH, and 17-OHP.

Four women with normal ovulation, two of whom previ-ously had endometriosis diagnosed and one of whom haduterine myomas, were selected as the control group becausethey had regular menses and no evidence of hirsutism. Thesewomen were matched by age and weight with the womenwith PCOS.

The mean (6SEM) age of the women with normal ovu-lation (30.56 3.1 years) was not significantly different thanthat of the women with PCOS (23.76 3 years). The mean(6SEM) body mass index of the women with normal ovu-lation was 31.86 4.9 kg/m2, whereas that of the womenwith PCOS was 306 3.1 kg/m2. All four control subjectsdemonstrated a biphasic BBT curve and a luteal phase pro-gesterone level of.6 ng/mL.

None of the subjects in either group had received anyhormonal medication for at least 6 weeks before the study.This study was approved by our institutional review board,and written informed consent was obtained from each subject.

ProtocolAll the study subjects received a GnRH-a (Lupron depot

7.5 mg IM; TAP Pharmaceutical Corp., Deerfield, IL)monthly at 8AM for 6 months. The injections in subjects withPCOS were begun without regard to the patient’s menstrualperiod and those in women with normal ovulation werebegun on day 3 of the menstrual cycle.

Each subject underwent a baseline evaluation beforeGnRH-a treatment that included determination of hormonelevels and assessment of adrenal dynamic responses afterACTH stimulation. In women with PCOS, blood sampleswere collected at 8AM every 3 days (27, 24, and21 days)and pooled to represent basal secretion. In women withnormal ovulation, blood samples for basal secretion werecollected at 8AM on day 2 of the menstrual cycle.

All the subjects received dexamethasone, 1 mg orally at11 PM on the evening that basal blood drawing was com-pleted and at 7AM the next morning. The protocol forphysiologic ACTH stimulation was as follows: SyntheticACTH (1-24, Cortrosyn; Organon Pharmaceuticals, WestOrange, NJ) was administered at a dose of 200 ng intrave-nously in two boluses at 9AM and 11AM, respectively. Thiswas followed by a more conventional pharmacologic stim-ulation with 0.25 mg of synthetic ACTH (1-24) intrave-nously at 1PM.

Blood samples were collected through an indwelling cath-eter at215, 0, 15, and 30 minutes after each physiologicACTH bolus and at215, 0, 60, and 120 minutes after thepharmacologic ACTH stimulation. At the completion of 6months of GnRH-a administration, basal secretion and theadrenal dynamic responses to ACTH were reevaluated in afashion identical to that of the pretreatment protocol.

440 Gonzalez et al. Physiologic vs. pharmacologic ACTH in PCOS Vol. 71, No. 3, March 1999

AssaysLuteinizing hormone, E2, progesterone, 17-OHP, testos-

terone, androstenedione, DHEA, DHEAS, and 11b-hy-droxyandrostenedione (11b-OHA) were measured with theuse of validated RIAs, by methods described previously(17–20). The intra-assay and interassay coefficients of vari-ation were,6% and 15%, respectively, for all assays.

Data AnalysisHormonal measurements obtained from blood drawn after

each physiologic ACTH bolus were averaged to representthe postphysiologic ACTH adrenal dynamic response. Ad-renal androgen sensitivity was defined as the maximumvalues of DHEA, androstenedione, and 11b-OHA achievedafter physiologic ACTH administration. The activities of17a-hydroxylase and 17,20-lyase were represented by themean ratios of the maximum values of 17-OHP to proges-terone and of androstenedione to 17-OHP, respectively, aftereither physiologic or pharmacologic ACTH stimulation.

Women with PCOS in whom the postpharmacologicACTH maximum value of androstenedione and the ratio ofthe maximum values of androstenedione to 17-OHP were.2 SD above the mean of those in women with normalovulation were considered to have 17,20-lyase hyperactivity(9). Only D4 steroid pairs and not those of theD5 pathwaywere used to evaluate enzyme activity based on previousdata (16) demonstrating an estrogen effect on 17,20-lyaseactivity primarily through theD4 pathway.

Statistical evaluation was performed with the use of Stu-dent’s paired and unpairedt-tests where appropriate. One-tailed significance was considered sufficient when pairedcomparisons were assessed for the basal serum levels of E2,testosterone, and androstenedione, because only declines inthese hormones were anticipated in response to GnRH-aadministration, as demonstrated previously (2, 21–23). How-ever, two-tailed significance was used for all other compar-isons. Data are expressed as means6 SEM.

RESULTS

Basal Hormonal ResponseThe mean (6SEM) baseline LH concentration was sig-

nificantly (P,.01) higher in the women with PCOS (14.462.5 mIU/mL) than in the women with normal ovulation(4.26 0.7 mIU/mL). Table 1 depicts the mean basal steroidhormone values in both groups before and after GnRH-atreatment. The women with PCOS exhibited significantly(P,.05) higher baseline levels of E2, testosterone, andro-stenedione, and DHEAS compared with those of the womenwith normal ovulation.

In the women with normal ovulation, the mean levels ofserum E2, androstenedione, and testosterone were signifi-cantly (P,.05) reduced, after GnRH-a treatment, whereasthose of serum DHEAS and 11b-OHA remained unaffected.In the women with PCOS, the mean levels of serum E2 andandrostenedione declined significantly (P,.05) afterGnRH-a treatment. The mean serum testosterone concentra-tion exhibited a moderate decline in women with PCOS andwas not statistically significant. In both study groups, serumE2 levels after GnRH-a treatment declined to,33 pg/mL.

Adrenal Dynamic Response Before GnRH-aAdministration

The physiologic ACTH responses of DHEA, andro-stenedione, and 11b-OHA before GnRH-a treatment, repre-senting baseline adrenal androgen sensitivity, are shown inFigure 1. Although there was no difference in the meanresponse of DHEA between the groups, the mean responsesof androstenedione and 11b-OHA were twofold higher inthe women with PCOS compared with the controls, whichwas statistically significant (P,.05).

The ratios of 17-OHP to progesterone and of andro-stenedione to 17-OHP in response to physiologic ACTHstimulation at the beginning of the study are depicted inFigure 2. Both these ratios were significantly (P,.05)

T A B L E 1

Mean basal steroid hormone levels of study subjects before and after GnRH agonist treatment.

Hormone

Patient group

Polycystic ovary syndrome Normal ovulation

Beforetreatment

Aftertreatment

Beforetreatment

Aftertreatment

E2 level (pg/mL) 70.56 7.0 33.06 8.0* 43.56 7.0† 26.86 6.0*Testosterone level (ng/dL) 56.36 7.7 45.96 7.4 37.46 3.4† 26.26 3.0*Androstenedione level (ng/mL) 4.36 0.7 2.46 0.8* 2.96 0.4† 2.16 0.2*DHEAS level (mg/dL) 4106 19 3806 64 1586 37† 1476 4111b-OHA level (ng/mL) 2.96 0.4 2.86 0.4 2.16 0.4 1.96 0.3

Note: All values are means6 SEM. 11b-OHA 5 11b-hydroxyandrostenedione.* P ,.05 (before GnRH agonist treatment vs. after GnRH agonist treatment in either group).† P ,.05 (PCOS group vs. control group).

FERTILITY & STERILITY t 441

greater in the women with PCOS compared with the con-trols.

As shown in Figure 3, there was no difference in the ratioof 17-OHP to progesterone between the groups in responseto pharmacologic ACTH stimulation at the beginning of thestudy. However, the ratio of androstenedione to 17-OHPafter pharmacologic ACTH stimulation was significantly (P,.05) higher in the women with PCOS compared with thewomen with normal ovulation before GnRH-a administra-tion. This higher pretreatment ratio of androstenedione to

17-OHP in response to pharmacologic ACTH stimulation,reflecting increased 17,20-lyase activity, was attributed tothree patients with PCOS (depicted by triangles in Fig. 3)who met the criteria for 17,20-lyase hyperactivity.

Adrenal Dynamic Response After GnRH-aadministration

After 6 months of GnRH-a administration, all responsesto physiologic ACTH stimulation were similar in the womenwith PCOS and the controls. For example, the previouslyelevated androstenedione and 11b-OHA responses in thewomen with PCOS (1.86 0.4 ng/mL and 1.46 0.4 ng/mL;respectively) were similar to those in the women with nor-mal ovulation (0.96 0.2 ng/mL and 0.86 0.1 ng/mL,respectively). The ratios of 17-OHP to progesterone and ofandrostenedione to 17-OHP also were similar in the twogroups after GnRH-a treatment (1.46 0.2 and 3.66 0.6,respectively, in the PCOS group and 1.36 0.1 and 2.760.8, respectively, in the control group).

In response to pharmacologic ACTH stimulation afterGnRH-a administration, the ratio of 17-OHP to progesteronein the women with PCOS remained unchanged comparedwith the pretreatment value, and the significantly (P,.05)elevated ratio of androstenedione to 17-OHP (1.66 0.4)persisted in this group compared with that of the controlgroup (0.86 0.1). All three of the women with PCOS whoexhibited 17,20-lyase hyperactivity before GnRH-a treat-ment continued to meet the criteria for this enzyme abnor-mality after GnRH-a treatment.

DISCUSSION

The women with PCOS in our study exhibited baselinehormonal characteristics that concur with previously pub-lished findings for this syndrome (2, 9, 24). Declines in thebasal serum levels of E2, androstenedione, and testosteronein the women with PCOS after GnRH-a administration in-dicate a reduction in ovarian steroidogenesis that is consis-tent with the results of several previous studies (2, 13, 14,16, 21, 22).

The baseline increase in adrenal androgen sensitivity afterthe administration of a physiologic dose of ACTH in womenwith PCOS is similar to our previous findings with the use ofcorticotropin-releasing hormone administration (4, 16). Toour knowledge, the significant baseline elevations in theactivities of 17a-hydroxylase and 17,20-lyase after physio-logic ACTH stimulation in these patients compared with thecontrols is a novel finding. In contrast, only the 17,20-lyasearm of cytochrome P450c 17a demonstrated a baseline in-crease in activity after pharmacologic ACTH administrationin the women with PCOS compared with the controls.

In this study, the 17,20-lyase hyperactivity was present ina subset (3 of 6) of these patients. This finding concurs withour initial description of this enzyme abnormality in a largerseries of women with PCOS who had elevated adrenal an-

F I G U R E 1

Baseline postphysiologic ACTH responses (mean 6 SEM)representing adrenal sensitivity for androstenedione (A; leftaxis) and 11b-hydroxyandrostenedione (11b-OHA; left axis)as well as DHEA (right axis) in women with normal ovulation(h) and women with polycystic ovary syndrome (PCOS; u). *Statistically significant difference (P,.05) between womenwith normal ovulation and women with PCOS.

F I G U R E 2

Baseline postphysiologic ACTH ratios (mean 6 SEM) of 17-hydroxyprogesterone (17-OHP) to progesterone (P) and ofandrostenedione (A) to 17-OHP in women with normal ovu-lation (h) and women with polycystic ovary syndrome(PCOS, u). * Statistically significant difference (P,.05) be-tween women with normal ovulation and women with PCOS.

442 Gonzalez et al. Physiologic vs. pharmacologic ACTH in PCOS Vol. 71, No. 3, March 1999

drogen levels (9). Because the hallmark of cytochromeP450c 17a dysregulation as described for the ovary byprevious investigators (10) is a relative increase in 17a-hydroxylase activity compared with that of 17,20-lyase, ourfindings are at variance with the concept of adrenal cyto-chrome P450c 17a dysregulation in women with PCOS.

The exaggerated adrenal responses to physiologic ACTHstimulation in our subjects with PCOS were observed tonormalize after suppression of ovarian steroidogenesis. Thiswas suggested by the lack of statistically significant differ-ences between the groups in the responses of androstene-dione and 11b-OHA as well as the ratios of 17-OHP toprogesterone and of androstenedione to 17-OHP after phys-iologic ACTH stimulation at the completion of GnRH-aadministration.

Although the number of patients in our pilot study is toosmall to draw a definitive conclusion, the possibility thatovarian function may influence adrenal androgen sensitivityand adrenal cytochrome P450c 17a activity in women withPCOS cannot be excluded. Significant declines in adrenalandrogen responses to standard doses of ACTH afterGnRH-a treatment have been shown previously in womenwith PCOS who have adrenal hyperandrogenism (13, 14).These data, coupled with those of our current study and ourprevious report (16), support the contention that estrogen inparticular may be the ovarian steroid responsible for induc-ing increased adrenal androgen sensitivity and increasedcytochrome P450c 17a activity in women with PCOS underphysiologic circumstances.

We found that the 17,20-lyase hyperactivity uncovered bypharmacologic ACTH stimulation in a subset of women withPCOS was not influenced by a reduction in ovarian steroids.This enzyme abnormality persisted in affected individualsduring this form of adrenal dynamic testing at the end ofGnRH-a treatment. A recently discovered point mutation incytochrome P450 17a believed to modify the phenotypicexpression of PCOS has been shown to upregulate the tran-scription of cytochrome P450 17a (25). On the other hand,excessive serine phosphorylation of cytochrome P450 17a toincrease cofactor binding has been proposed as a mechanismfor promoting 17,20-lyase hyperactivity (26).

Thus, 17,20-lyase hyperactivity probably is an intrinsicadrenal abnormality and may be the result of a genetic orregulatory defect that is evident in some, but not all, womenwith PCOS who have adrenal androgen excess. However,this adrenal enzyme defect does not appear to be modified byestrogen or other ovarian factors.

In summary, we identified separate adrenal abnormalitiesin response to physiologic and pharmacologic ACTH stim-ulation in women with PCOS who had elevated adrenalandrogen levels. Increased adrenal androgen sensitivity andan increase in the activities of both arms of cytochromeP450c 17a occurred after physiologic ACTH stimulationthat may have resolved after GnRH-a induced modulation ofovarian steroids. In contrast, the 17,20-lyase hyperactivitythat was evident in some patients after pharmacologic ACTHstimulation persisted after long-term GnRH-a treatment.

F I G U R E 3

Baseline postpharmacologic ACTH ratios (●) of 17-hydroxyprogesterone (17-OHP) to progesterone (P) and of androstenedione(A) to 17-OHP as well as baseline maximum absolute values of A in women with normal ovulation and women with polycysticovary syndrome (PCOS). The horizontal lines and error bars represent the mean 6 SEM of values for each group and the boxesrepresent 2 SD above and below the mean of values in women with normal ovulation. Three women with PCOS (�) met thecriteria for 17,20-lyase hyperactivity. * Statistically significant difference (P,.05) between women with normal ovulation andwomen with PCOS.

FERTILITY & STERILITY t 443

In conclusion, we propose that in women with PCOS,physiologic ACTH stimulation may uncover acquired adre-nal defects promoted by the ovary in the anovulatory state,whereas pharmacologic ACTH stimulation is necessary toexpose a subtle intrinsic adrenal abnormality that appears tobe independent of an ovarian influence.

Acknowledgments:We thank TAP Pharmaceuticals, Deerfield, Illinois, fordonating the Lupron depot used in this study. We also thank CarmenTodaro, M.D., Rosendo Intengan, M.D., Mario Reyes, M.D., Judith Ort-man-Nabi, M.D., and Sheri Baczkowski, M.D., for their prompt patientreferrals from their respective gynecologic practices in Buffalo, New York,and Miguel Rodriguez-Bigas, M.D., of the State University of New York atBuffalo for the thorough statistical evaluation.

References1. Gonzalez F. Adrenal involvement in polycystic ovary syndrome. Semin

Reprod Endocrinol 1997;15:137–59.2. Gonzalez F, Hatala DA, Speroff L. Adrenal and ovarian steroid hor-

mone responses to gonadotropin releasing hormone agonist treatment inwomen with polycystic ovary syndrome. Am J Obstet Gynecol 1991;195:535–45.

3. Hoffman DI, Klove K, Lobo RA. The prevalence and significance ofelevated dehydroepiandrosterone sulfate levels in anovulatory women.Fertil Steril 1984;42:76–81.

4. Carmina E, Lobo RA. Pituitary-adrenal responses to ovine corticotrop-in-releasing factor in polycystic ovary syndrome and in other hyperan-drogenic patients. Gynecol Endocrinol 1990;4:225–32.

5. Gonzalez F, Speroff L. Adrenal morphologic considerations in poly-cystic ovary syndrome. Obstet Gynecol Surv 1990;45:491–508.

6. Chang RJ, Mandel FP, Wolfsen AR, Judd HL. Circulating levels ofplasma adrenocorticotropin in polycystic ovarian disease. J Clin Endo-crinol Metab 1982;54:1265–67.

7. Lobo RA, Goebelsmann U. Adult manifestations of congenital adrenalhyperplasia due to incomplete 21-hydroxylase deficiency mimickingpolycystic ovarian disease. Am J Obstet Gynecol 1980;138:720–4.

8. Azziz R, Bradley EL, Potter HD, Boots LR. 3b-Hydroxysteroid dehy-drogenase deficiency in hyperandrogenism. Am J Obstet Gynecol 1993;168:889–95.

9. Gonzalez F, Chang L, Horab T, Lobo RA. Evidence for heterogeneousetiologies of adrenal dysfunction in polycystic ovary syndrome. FertilSteril 1996;66:354–61.

10. Barnes RB, Rosenfield RL, Burstein S, Ehrmann DA. Pituitary-ovarianresponses to nafarelin testing in the polycystic ovary syndrome. N EnglJ Med 1989;320:559–65.

11. Rosenfield RL, Barnes RB, Cara JF, Lucky AW. Dysregulation ofcytochrome P450 17a as the cause of polycystic ovary syndrome. FertilSteril 1990;53:785–91.

12. Azziz R, Bradley EL, Potter HD, Boots LR. Adrenal androgen excessin women: lack of a role for 17-hydroxylase and 17,20 lyase dysregu-lation. J Clin Endocrinol Metab 1995;80:400–5.

13. Carmina E, Gonzalez F, Chang L, Lobo RA. Reassessment of adrenalandrogen secretion in women with polycystic ovary syndrome. ObstetGynecol 1995;85:971–6.

14. Fruzzetti F, DeLorenzo D, Ricci C, Teti G. Ovarian influence onadrenal androgen secretion in polycystic ovary syndrome. Fertil Steril1995;63:734–41.

15. Mattieri RK, Stanczyk FZ, Cassidente DL, Paulson RJ, Lobo RA. Theovarian contribution to peripherally derived C19 conjugates. J ClinEndocrinol Metab 1992;75:768–72.

16. Ditkoff EC, Fruzzetti F, Chang L, Stanczyk FZ, Lobo RA. The impactof estrogen on adrenal androgen sensitivity and secretion in polycysticovary syndrome. J Clin Endocrinol Metab 1995;80:603–7.

17. Lobo RA, Kletsky OA, Kaptein EM, Goebelsmann U. Prolactin mod-eration of dehydroepiandrostenedione sulfate secretion. Am J ObstetGynecol 1980;138:632–6.

18. Stanczyk FZ, Chang L, Carmina E, Putz Z, Lobo RA. Is 11b-hydroxy-androstenedione a better marker of adrenal androgen excess than de-hydroepiandrostenedione sulfate? Am J Obstet Gynecol 1991;165:1837–42.

19. Mishell DR Jr, Nakamura RM, Crosignani PG, Stone S, Kharma K,Nagata Y, et al. Serum gonadotropin and steroid patterns during themenstrual cycle. Am J Obstet Gynecol 1971;119:445–52.

20. Goebelsman U, Arce JJ, Thorneycroft IH, Mishell DR Jr. Serumtestosterone concentrations in women throughout the menstrual cycleand following HCG administration. Am J Obstet Gynecol 1974;119:445–52.

21. Chang RJ, Laufer LR, Meldrum DR, DeFazio J, Lu JKH, Vale W, et al.Steroid secretion in polycystic ovarian disease after ovarian suppressionby a long-acting gonadotropin releasing hormone agonist. J Clin En-docrinol Metab 1983;56:897–903.

22. Steingold, DeZiegler D, Cedars M, Lu JKH, Meldrum DR, Judd HL, etal. Clinical and hormonal effects of chronic gonadotropin-releasinghormone agonist treatment in polycystic ovarian disease. J Clin Endo-crinol Metab 1986;65:773–8.

23. Rittmaster RS. Differential suppression of testosterone and estradiol inhirsute women with the superactive gonadotropin-releasing hormoneagonist leuprolide. J Clin Endocrinol Metab 1988;67:651–5.

24. DeVane GW, Czekula NM, Judd HL, Yen SSC. Circulating gonado-tropins, estrogens and androgens in polycystic ovarian disease. Am JObstet Gynecol 1975;121:496–500.

25. Carey AH, Waterworth D, Patel K, White D, Little J, Novelli P, et al.Polycystic ovaries and male pattern baldness are associated with oneallele of the steroid metabolism gene CYP17. Hum Mol Genet 1994;3:1873–6.

26. Zhang L, Rodriguez H, Ohon S, Miller WL. Serine phosphorylation ofhuman P450 c17 increases 17,20 lyase activity: implications for adre-narche and for the polycystic ovary syndrome. Proc Natl Acad Sci USA1995;92:10619–23.

444 Gonzalez et al. Physiologic vs. pharmacologic ACTH in PCOS Vol. 71, No. 3, March 1999