- Email: [email protected]
A practical dexamethasone suppression test to evaluate hirsute women
Int. J. Gynaecol. Obstet., 1982,20: 35-41 International Federation of Gynaecology & Obstetrics
A PRACTICAL DEXAMETHASONE SUPPRESSION TEST TO EVALUATE HIRSUTE WOMEN
CHUNG H. WU Division of Reproductive Endocrinology, Department of Thomas Jefferson University, Philadelphia, PA 19107, USA
Obstetrics and
Gynecology, Jefferson
Medical College,
(Received June 22nd, 1981) (Accepted July 29th, 1981)
Abstract Wu CH (Division of Reproductive Endocrinology, Dept of Obstetrics and Gynecology, Jefferson Medical College, Thomas Jefferson University, Philadelphia, PA I91 0 7, USA). A practical dexamethasone suppression test to evaluate hirsute women. Int J Gynaecol Obstet 20: 35-41, 1982 Fifty five hirsute women were subjected to a 2-week dexamethasone (DXMI suppression test. The pre- and post-DXM plasma dehydroepiandrosterone-sulfate (OS) and testosterone (T) were measured by radioimmunoassay to define the source of androgen excess in hirsute women. Four patients (7%) failed to have adequate adrenal suppression due to failure in medication. Among the 51 patients with adequate adrenal suppression, the source of androgen excess was clearly defined in 48 patients (94%). Seventeen patients (33%) showed ovarian source, 13 patients (26%) had adrenal source, while 18 patients (35%) revealed a mixed adrenal and ovarian source. Normal baseline DS and T levels were noted in 22% of hirsute women and more than half (55%) of them had ovarian androgen excess. Even in 17 patients with normal DS and elevated T, 6 patients (36%) suggested adrenal androgen excess, The source of androgen excess in hirsute women seems evenly distributed among the ovarian, the adrenal, and the mixed group.
Key words: Dexamethasone
suppression test;
002-7292/82/0000-0000/$02.75 0 1982 International Federation of Gynaecology & Obstetrics
DXM plasma dehydroepiandrosterone-sulfate; Testosterone; Radioimmunoassay; Adrenal suppression; Hirsutism; Venous catheterization of adrenal and ovarian veins. Introduction Hirsutism is a common and disturbing problem frequently associated with menstrual disorder. Androgen excess has been proved in almost every hirsute woman [ 61. However, the source of androgen excess in hirsute women is still controversial. Using venous catheterization of adrenal and ovarian veins, Kirschner [ 11 ,121 showed that most of the androgen excess in hirsute women is mainly ovarian origin. In contrast, on the basis of dexamethasone (DXM) suppression test, Abraham 141 showed that most of the androgen excess in hirsute women, is adrenal origin. The former method requires surgical procedure and is neither harmless, nor practical. The latter method requires measurement of five major androgens and cortisol before and after DXM suppression. This approach is certainly more feasible. However, the quantitation of five different androgens is not only expensive but also non-obtainable in the average clinical laboratory. Therefore, there is a need to find a relatively simple laboratory test to define the androgen source in hirsute women. Clinically, it is important to define the source of androgen excess because it will dictate the subsequent therapy. The following study was carried out by measuring plasma dehydroepiandrosteroneInt J Gynaecol Obstet 20
36
wu
sulfate (DS) and testosterone (T) before and after 2 weeks of DXM medication in hirsute women. Evaluating the differences between the pre- and the post-DXM DS and T level, it was possible to define the source of androgen excess in hirsute women.
Materials and methods A series of 55 consecutive women with chief complaint of “excess hair growth” were enrolled in this study after initial medical history, physical examination and the basic endocrine work-up. The dosage of DXM was varied according to the body weight as recommended by Abraham [4]. Blood samples were obtained before and after DXM medication and the plasma were separated and stored at -20°C until assayed. The plasma DS and T were measured by respective radioimmunoassays [ 8,201. Dexamethasone was administered for 2 weeks period as adrenal androgen secretion requires approximately 2 weeks of continuous DXM suppression to reach a basal level [ 1,4]. Normal upper limit (mean + 2 S.D.) for DS in normal females (control: 1.61 rt 0.12 (mean + S.E.) fig/ml, y1=14) is 2.5 pg/ml and the post-DXM DS level, less than 0.5 pg/ml, was considered as adequate adrenal suppression [ 1.5I. Those patients with pre-DXM DS level higher than 2.5 pg/ml is considered as “adrenal androgen Table I
excess”. Those patients with the post-DXM DS level higher than 0.5 pg/ml were considered “incomplete suppression”, and their data were deleted without analysis. Only four patients failed to be adequately suppressed during the study period. They all failed to take DXM as instructed. Approximately the ovary contributes 40-45% of T in peripheral circulation while the adrenal contributes approximately 55-60% [ 1,21 I. The upper normal limit (mean + 2 S.D.) of T in normal females (control: 42 f 2.4 (mean f S.E.) ng/dl, n= 14) was 60 ng/dl in our laboratory. Therefore, approximately 35 ng/dl of circulating plasma T is contributed by the adrenal, where the rest of 25 ng/dl is contributed by the ovary. A difference of the pre- and post-DXM T level less than 35 ng/dl is considered a normal adrenal T secretion. Thus if the difference is larger than 35 ng/dl, it is considered “adrenal androgen excess”. The post-DXM T level in the normal females should be less than 25 ng/dl; thus if the postDXM T level higher than 25 ng/dl, it is considered “ovarian androgen excess”. In addition to the hormonal data, the menstrual history and clinical signs of androgen excess such as acne, male escatcheon, and hirsutism were recorded and analysed. Polycystic ovarian disease (PCO) was diagnosed by laparoscopic examination, elevated LH level and clinical signs. Statistical analysis was carried out by the Student’s t-test.
Pre- and Post- dexamethasone (DXM) suppression plasma DHA - sulfate and testosterone in hirsute females
source of androgen excess
Ovarian Adrenal Mixed None All patients
No. (%)
17(33) 13(26) 18(35) 3(6) Sl(100)
Testosteronea ~----_.-
DHA-sulfatea Re-DXM
PostDXM
PX-DXM
Post-DXM
1.71 2.91 3.54 1.47 2.65
0.31 0.25 0.33 0.14 0.29
75 74 81 51 75
65 16 49 22 43
f 0.11u f 0.34d + 0.25d f 0.36 f 0.17d
f f f f f
o.02c o.03c o.02c o.03c o.02c
f 7.2d f 4.9d f 6.3d f 3.8 * 3.6d
aUpper normal limits of pre-DXM DS = 2.5 Mg/ml, T = 60 ng/dL While upper normal Limitsof post-DXM DS = 0.5 rg/ml, T = 25 ng/dl. bMean f S.E. ‘P < 0.0 1 when compared with the pm-DXM value. dP < 0.01 when compared with the normal follicular phase levels (DHAS: 1.61 f 0.12 &ml, T: 42 + 2.4 ngldl, u = 14). ZntJ Gynaecol Obstet 20
f 7.6 f 1.5’ f 3.6’ f 1.2c &4.oc
DXM suppression teest in hirsu tism
Results Fifty-one out of 55 patients who had gone through the DXM suppression test showed an adequate adrenal suppression (i.e. post-DXM DS less than 0.5 pg/ml). The data of the four patients with incomplete adrenal suppression (i.e. post-DXM DS higher than 0.5 E.cg/ml) were deleted. Among these fifty-one patients, 17 (33%) patients showed androgen excess of ovarian source (ovarian group), 13 (26%) patients had adrenal source (adrenal group), while 18 (35%) patients had a mixed ovarian and adrenal source (mixed group) of androgen Table II.
3I
excess. Three patients (6%) out of 5 1 patients failed to identify the source (none group) of androgen excess (Table I). Therefore, the source of androgen excess was identified by this DXM suppression test in 94% (48 out of 51) of hirsute women. Elevated pre-DXM DS mean level was noted in the hirsute women and this was mainly attributed by the elevated level in the adrenal and mixed group. In the group of ovarian androgen excess, the DS level was within normal range. The preDXM T level was elevated in all groups except with unidentifiable source the patients (none group) of androgen excess. Post-DXM T
Distribution of the clinical findings in hirsute females,
Clinical fmding
All patients
Source of androgen excess ovarian
Adrenal
Mixed __-_!-
None
51a(100)b (lOOjC
17(100) (33)
13(100) (26)
18(100) (35)
3(100) (6)
9(18) (100) 28(55) (100) 14(27) (100)
3(18) (33) lO(59) (36) 4(23) (2%
4(31) (45) 4(31) (14) 5(38) (35)
2(11) (22) 12(67) (43) 4(22) (29)
0) (0) 2(67) (7) l(33) (7)
29(57) (100) 15(29) (100) 7(14) (100)
8(47) (27) 3(18) (20) 6(35) (86)
8(61) (27) 4(31) (27) l(8) (14)
12(67) (42) 6(33) (40) O(O) (0)
I(331 (4) 2(67) (13) O(O) (0)
Acne
16(31) (100)
2(12) (13)
6(46) (37)
7(39) (44)
I(331 (6)
Male escatcheon
16(31) (100)
7(41) (44)
3(23) (19)
6(33) (37)
O(O) (0)
4(8) (100)
2(12) (50)
O(O) (0)
2(11) (50)
O(O) (0)
13(26) (100)
8(47) (62)
O(O) (0)
5(28) (38)
O(O) (0)
-..(All patients)
Menstrual pattern: Eumenorrhea Oligomenorrhea Amenorrhea
Hirsutism: Mild Moderate Severe
Clitoromegaly
Polycystic ovaries
aNumber of patient. b% in the vertical group. ‘% in the horizontal group. Int J Gynaecol Obstet 20
38
Wu
level was suppressed in all groups of patients except the ovarian group. In order to evaluate the relationship between the menstrual patterns and the clinical signs of androgen excess vs. the sources of androgen excess, the percent distribution of all the clinical signs were tabulated according to the source of androgen excess (Table II). Eighteen percent of total patients had eumenorrhea but oligomenorrhea was the most common menstrual pattern observed in hirsute women. Mild to moderate hirsutism was a common occurrence in these patients and only 31% of patients showed evidence in acne. Male escatcheon was observed in 16 patients (3 1%), whereas clitoromegaly was only observed in four patients (8%). Clinical diagnosis of polycystic ovarian disease was noted in 26% of the patients. When these clinical findings were distributed accordingly to their source of androgen excess, eumenorrhea was common in the adrenal group and oligomenorrhea-amenorrhea was more frequently observed in the ovarian or mixed groups. Severe hirsutism was quite common in the ovarian group, whereas mild hirsutism is more frequently seen in the adrenal and the mixed groups. Acne was less Table III.
common in the ovarian group, and male escatcheon associated less frequently with the adrenal group. Polycystic ovary was not seen in the adrenal group, yet common in the ovarian and mixed groups. The baseline DS and T level in this group of hirsute women was reevaluated retrospectively to see the frequency of abnormalities in the different groups of androgen excess (Table III). Nomzal DS and T Zevel was noted in 11 patients (22%) out of total 5 1 hirsute women and more than half (55%) of them belonged to the ovarian group. Only 7 patients had elevated DS and normal T level and most of them belonged to the mixed group and none in the ovarian group. Approximately one third of the patients showed normal DS and elevated T. Most of them (64%) were in the ovarian group. Approximately another one third (3 1%) of patients had both DS and T elevated. This latter combination was more frequently observed in the mixed group and none in the ovarian group. Even an elevated DS and/or T was considered, only 78% of hirsute women showed an abnormality, although, 100% of mixed group and 85% of adrenal group had abnormalities in this category.
Distribution of the baseline DHA-sulfate (DS) and testosterone (T) levels in hirsute females.
source of androgen excess
Androgen level DS: -a T: -
DS: tb T: -
DS: T: t
DS: t T: t
DS: f and/or T: t
llc(lOO)d we 6 (55) (35) 2(18) (15) O(0) (0) 3(27) (100)
7(100) (14) O(0) (0) X29) (15) 5(71) (28) O(0) (0)
17(100) (33) 1 l(64) (65) 4(24) (31) 202) (11) O(0) (0)
16(100) (31) O(0) (0) 5(31) (39) 1 l(69) (61) O(0) (0)
40(100) (78) 1 l(28) (65) 1 l(28) (85) 18(44) (100) O(0) (0)
-All patients n=Sl Ovarian n-17 Adrenal n=13 Mixed n=18 None n=3 a
-,
normal.
bt; elevated. %mber of patient. + Din the vertical group. %/oin the horizontal group. In? J Gynaecol Obstet 20
DXMsuppression
Discussion Dexamethasone suppression tests have been utilized to search for the source of androgen excess in hirsute women for the last couple of decades. Its value has been questioned or affiied [4,10]. One of the major problems encountered in previous DXM suppression tests was the inconsistancy of urinary 17ketosteroids in reflecting the androgen state. 17-Ketosteroids are androgen metabolites derived from adrenal as well as ovarian androgen; thus it does not exclusively represent the androgen activity of either gland [ 10,141. In addition, the dosage of DXM and the duration of medication were arbitrarily chosen without evaluating its adequacy. Thus interpretation of results can be clouded by data obtained under incomplete adrenal suppression. More than 90% of circulating DS is originated from the adrenal [ 1,3]. Its peripheral concentration is stable and its metabolic half-life is much longer [ 13,191, therefore, DS fluctuates less than cortisol under acute stress and it reflects better the adrenal androgen secretory activity. Although DS is ACTH dependent but its response to ACTH is not as rapid nor sensitive as corticosteroids [ 191. Perhaps adrenal androgen stimulating hormone (AASH), a possible new hormone has not been identified, or prolactin may stimulate the adrenal androgen secretion [ 7,191. During the DXM administration, cortisol level was suppressed to the baseline level within 2 or 3 days, yet DS would not reach a basal level during this period of time. Instead it takes approximately 2 weeks for DS to reach the baseline level [ 1,101. Therefore, it is advisable to administer DXM for 2 weeks in order to have a completely suppressed adrenal androgen secretion. Incomplete adrenal suppression occurring in a 2-day suppression test may complicate the interpretation of the result. Even in a 1 week DXM suppression, there were only 54% of patients who showed an adequate adrenal suppression [4]. In addition, the dosage of DXM should be varied
test in hirsutism
39
according to the body weight as suggested by Abraham [4]. This is especially true in hirsute women as they tend to have excessive weight gain. The sources of excess androgen in hirsute women was evaluated by measuring plasma concentration of androgens in the adrenal and ovarian vein blood [ 11,121. Its result suggested that in more than 90% of hirsute women their androgen excess was originated from the ovary. This elegant study, however, was questioned for its result because of the subsequent evidence disputed their theoretical assumption of a parallel secretory activity between adrenal androgen and cortisol [2]. Furthermore, the episodic secretory pattern of adrenal steroids [ 101 makes it difficult to interpretate the long-term adrenal function on the basis of data in venous blood samples obtained by catheterization. It’s finding that DXM suppresses ovarian androgen secretion was also questioned, because the androgen level in long-term DXM treated women was not different from those levels in the adrenalectomized women [ 1,3]. Therefore, DXM can be properly used to suppress adrenal androgen secretion in order to define the source of androgen excess in hirsute women. Adrenal androgen excess state may lead to a disrupted ovarian folliculogenesis and subsequently increases ovarian androgen secretion. Therefore, it is quite conceivable that initial adrenal androgen hyper-secretion may lead into a mixed ovarian and adrenal hypersecretion. The possibility of opposite phenomenon, i.e. initial ovarian androgen excess induces subsequent adrenal androgen excess, is not clear at the present time. However, in those patients with mixed androgen excess, it is possible to reverse ovarian androgen excess state, especially when patient becomes ovulatory. Because of this possibility of changing the source of androgen excess, it should be flexible in treating hirsute women and reevaluation may be necessary from time to time. A partial 1 I- and 21-hydroxylase deficiency was reported in a group of hirsute women with adrenal androgen excess [ 161. Iat J Gynaecol Obstet 20
40
wu
Table N.
Sources of androgen excess in hirsute women.
References
Patient No.
Kirschner and Jacobs [ 11 ] Givens et al. [9] Kirschner et al. [ 121 Toaff et al. [18] Abraham et al. [5] Wu (present study) TOTAL
Ovarian
Adrenal
Mixed
13 20 44 35 86 48
9/l 3(69%) 8/20(40%) 42/44(95%) 2/35(6%) 16/86(19%) 17/48(35%)
o/ 13(0%) 2/20(10%) 2/44(5%) 25/35(71%) 44/86(5 1%) 13(48(27%)
4/13(31%) 10/20(50%) O/44(0%) 8/35(23%) 26/86(30%) 18/48(38%)
246
94/246(38%)
86/246(35%)
66/246(27%)
Thus hirsute women should be carefully evaluated to rule out the mild or “attenuated” congenital adrenal hyperplasia [ 171. Oligomenorrhea and/or amenorrhea were the most common menstrual patterns in hirsute women. Eumenorrhea was more common in women with adrenal androgen excess. Perhaps the latter group of patient may still preserve ovulatory function, inspite of androgen excess on some occasions. Thus a mild adrenal hyper-androgen secretion may not disrupt ovulation. The patients with ovarian androgen excess tend to develop severe hirsutism. This might be related to the elevated T which was more frequently observed in this group of patients. Acne was seen in 31% of patients and most of them suggested adrenal origin of androgen excess. Thus, adrenal androgen is probably related to acne formation in hirsute women. Male escatcheon was evenly observed among the three groups and clitoromegaly was only observed in four patients thus there were too few patients to evaluate. Polycystic ovary was noted in 26% of patients. This low incidence observed in this study might be due to some patients who did not have laparoscopic examination or LH determination. It is possible that a higher incidence of polycystic ovary may occur in hirsute women. Source of androgen excess in hirsutism is among the probably evenly distributed adrenal, ovarian, and mixed groups (Table IV). Since the mixed group can be included in either the adrenal or the ovarian origin, it is Int J Gynaecol
Obstet 20
Source of androgen excess
quite acceptable to consider either the ovary or the adrenal as the major source of androgen excess in hirsutism. This study showed that measurement of DS and T before and after the 2-week DXM suppression can identify the source of androgen excess in 94% of hirsute women. This can be a simple and practical test in evaluation of hirsute women. References Abraham GE: Ovarian and adrenal contribution to peripheral androgens during the menstrual cycle. J CJin Endocrinol Metab 39: 340,1974. Abraham GE: Ovarian and adrenal contribution to peripheral steroids during the menstrual cycle in two hirsute women. Obstet Gynecol. 46: 29,1975. Abraham GE, Chakmakjian ZH: Serum steroid levels during the menstrual cycle in a bilaterally adrenalectomized woman. J Clin Endocrinol Metab 27: 581, 1973. Abraham GE, Maroulis GB, Buster JE et al: Effect of dexamethasone on serum cortisol and androgen levels in hirsute patients. Obstet Gynecol47: 395,1976. Abraham GE, Maroulis GB, Boyers SP et al: Dexamethasone suppression test in the management of hyperandrogenized patients. Obstet Gynecol5 7: 158,198 1. Bardin CW, Lipsett MB: Testosterone and androstenedione blood production rates in normal women and women with idiopathic hirsutism or polycystic ovaries. J Clin Endocrinol Metab 46: 891,1967. Bassi F, Giusti G, Borsi L et al: Plasma androgens in women with hyperprolactinaemic amenorrhea. Clin Endrocrinol6: 5,1977. Buster JE, Abraham GE: Radioimmunoassay for dehydroepiandrosterone sulfate. Anal Lett 5: 543: 1972. Givens JR, Andersen RN, Ragland JB et al: Andrenal function in hirsutism I. Diurnal change and response of plasma androstenedione, testosterone, 17-hydroxyprogesterone, cortisol, LH and FSH to dexamethasone and
DXM suppression test in him tism
10
11
12
13
14
15
16
yZ unit of ACTH. J Clin Endocrinol Metab 40: 988, 1975. Judd HL, McPherson RA, Rakoff JS et al: Correlation of the effects of dexamethasone administration on urinary 17ketosteroids and serum androgen levels in patients with hirsutism. Am J Obstet Gynecol 1.28: 408,1977. Kirschner MA, Jacobs JB: Combined ovarian and adrenal vein catheterization to determine the site(s) of androgen overproduction in hirsute women. J Clin Endocrinol Metab 33: 199,197l. Kirschner MA, Zucker IR, Jespersen, D: Idiopathic hirsutism - an ovarian abnormality. N Engl J Med 294: 637,1976. KorthSchutz S, Levine LS, New MI: Dehydroepiandrosterone sulfate (DS) levels, a rapid test for abnormal adrenal androgen secretion. J Clin Endocrinol Metab 42: 1005,1976. Lobo RA, Paul W, Goebelsmann U: Clinical use of plasma DHEA-sulfate: correlations with total and fractionated urinary 17-ketosteroids. The pacific mast Fertility Society Meeting 1979. Abstract p. 27. Maroulis GB, Manlimos FS, Abraham GE: Comparison between urinary 17ketosteroids and serum androgens in hirsute patients. Obstet Gynecol49: 454,1977. Newmark S, Dluhy RG, Williams GH et al: Partial lland 21-hydroxylase deficiencies in hirsute women. Am J Obstet Gynecoll27: 594, 1977.
41
17 Rosenwaks Z, Lee PA, Jones GS et al: An attenuated form of congenital adrenal hyperplasia. J Clin EndocrinolMetab49: 335,1979. 18 Toaff, R, Toaff ME, Gould S et al: Role of androgenie hyperactivity in anovulation. Fertil Steril 29: 407,1978. 19 Vaitukaitis JL, Cale SL, Melby JC: Role of ACTH in the secretion of free dehydroepiandrosterone and its sulfate ester in man. J Clin Endocrinol Metab 29: 1443,1969. 20 Wu CH, Blasco L, Flickinger GL et al: Ovarian function in the preovulatory rabbit. Biol Reprod 17: 304, 1977. 21 Yen SSC, Jaffe RB: Reproductive Endocrinology, physiology, pathophysiology and clinical management, p. 301. W.B. Saunders Company, 1978.
Address for reprints:
Chung H. Wu, M.D. Dept of Obstetrics and Gynecology Jefferson Medical College Room 300 1025 Walnut Street Philadelphia, PA 19107 USA
Znt J Gynaecol Obstet 20
A PRACTICAL DEXAMETHASONE SUPPRESSION TEST TO EVALUATE HIRSUTE WOMEN
CHUNG H. WU Division of Reproductive Endocrinology, Department of Thomas Jefferson University, Philadelphia, PA 19107, USA
Obstetrics and
Gynecology, Jefferson
Medical College,
(Received June 22nd, 1981) (Accepted July 29th, 1981)
Abstract Wu CH (Division of Reproductive Endocrinology, Dept of Obstetrics and Gynecology, Jefferson Medical College, Thomas Jefferson University, Philadelphia, PA I91 0 7, USA). A practical dexamethasone suppression test to evaluate hirsute women. Int J Gynaecol Obstet 20: 35-41, 1982 Fifty five hirsute women were subjected to a 2-week dexamethasone (DXMI suppression test. The pre- and post-DXM plasma dehydroepiandrosterone-sulfate (OS) and testosterone (T) were measured by radioimmunoassay to define the source of androgen excess in hirsute women. Four patients (7%) failed to have adequate adrenal suppression due to failure in medication. Among the 51 patients with adequate adrenal suppression, the source of androgen excess was clearly defined in 48 patients (94%). Seventeen patients (33%) showed ovarian source, 13 patients (26%) had adrenal source, while 18 patients (35%) revealed a mixed adrenal and ovarian source. Normal baseline DS and T levels were noted in 22% of hirsute women and more than half (55%) of them had ovarian androgen excess. Even in 17 patients with normal DS and elevated T, 6 patients (36%) suggested adrenal androgen excess, The source of androgen excess in hirsute women seems evenly distributed among the ovarian, the adrenal, and the mixed group.
Key words: Dexamethasone
suppression test;
002-7292/82/0000-0000/$02.75 0 1982 International Federation of Gynaecology & Obstetrics
DXM plasma dehydroepiandrosterone-sulfate; Testosterone; Radioimmunoassay; Adrenal suppression; Hirsutism; Venous catheterization of adrenal and ovarian veins. Introduction Hirsutism is a common and disturbing problem frequently associated with menstrual disorder. Androgen excess has been proved in almost every hirsute woman [ 61. However, the source of androgen excess in hirsute women is still controversial. Using venous catheterization of adrenal and ovarian veins, Kirschner [ 11 ,121 showed that most of the androgen excess in hirsute women is mainly ovarian origin. In contrast, on the basis of dexamethasone (DXM) suppression test, Abraham 141 showed that most of the androgen excess in hirsute women, is adrenal origin. The former method requires surgical procedure and is neither harmless, nor practical. The latter method requires measurement of five major androgens and cortisol before and after DXM suppression. This approach is certainly more feasible. However, the quantitation of five different androgens is not only expensive but also non-obtainable in the average clinical laboratory. Therefore, there is a need to find a relatively simple laboratory test to define the androgen source in hirsute women. Clinically, it is important to define the source of androgen excess because it will dictate the subsequent therapy. The following study was carried out by measuring plasma dehydroepiandrosteroneInt J Gynaecol Obstet 20
36
wu
sulfate (DS) and testosterone (T) before and after 2 weeks of DXM medication in hirsute women. Evaluating the differences between the pre- and the post-DXM DS and T level, it was possible to define the source of androgen excess in hirsute women.
Materials and methods A series of 55 consecutive women with chief complaint of “excess hair growth” were enrolled in this study after initial medical history, physical examination and the basic endocrine work-up. The dosage of DXM was varied according to the body weight as recommended by Abraham [4]. Blood samples were obtained before and after DXM medication and the plasma were separated and stored at -20°C until assayed. The plasma DS and T were measured by respective radioimmunoassays [ 8,201. Dexamethasone was administered for 2 weeks period as adrenal androgen secretion requires approximately 2 weeks of continuous DXM suppression to reach a basal level [ 1,4]. Normal upper limit (mean + 2 S.D.) for DS in normal females (control: 1.61 rt 0.12 (mean + S.E.) fig/ml, y1=14) is 2.5 pg/ml and the post-DXM DS level, less than 0.5 pg/ml, was considered as adequate adrenal suppression [ 1.5I. Those patients with pre-DXM DS level higher than 2.5 pg/ml is considered as “adrenal androgen Table I
excess”. Those patients with the post-DXM DS level higher than 0.5 pg/ml were considered “incomplete suppression”, and their data were deleted without analysis. Only four patients failed to be adequately suppressed during the study period. They all failed to take DXM as instructed. Approximately the ovary contributes 40-45% of T in peripheral circulation while the adrenal contributes approximately 55-60% [ 1,21 I. The upper normal limit (mean + 2 S.D.) of T in normal females (control: 42 f 2.4 (mean f S.E.) ng/dl, n= 14) was 60 ng/dl in our laboratory. Therefore, approximately 35 ng/dl of circulating plasma T is contributed by the adrenal, where the rest of 25 ng/dl is contributed by the ovary. A difference of the pre- and post-DXM T level less than 35 ng/dl is considered a normal adrenal T secretion. Thus if the difference is larger than 35 ng/dl, it is considered “adrenal androgen excess”. The post-DXM T level in the normal females should be less than 25 ng/dl; thus if the postDXM T level higher than 25 ng/dl, it is considered “ovarian androgen excess”. In addition to the hormonal data, the menstrual history and clinical signs of androgen excess such as acne, male escatcheon, and hirsutism were recorded and analysed. Polycystic ovarian disease (PCO) was diagnosed by laparoscopic examination, elevated LH level and clinical signs. Statistical analysis was carried out by the Student’s t-test.
Pre- and Post- dexamethasone (DXM) suppression plasma DHA - sulfate and testosterone in hirsute females
source of androgen excess
Ovarian Adrenal Mixed None All patients
No. (%)
17(33) 13(26) 18(35) 3(6) Sl(100)
Testosteronea ~----_.-
DHA-sulfatea Re-DXM
PostDXM
PX-DXM
Post-DXM
1.71 2.91 3.54 1.47 2.65
0.31 0.25 0.33 0.14 0.29
75 74 81 51 75
65 16 49 22 43
f 0.11u f 0.34d + 0.25d f 0.36 f 0.17d
f f f f f
o.02c o.03c o.02c o.03c o.02c
f 7.2d f 4.9d f 6.3d f 3.8 * 3.6d
aUpper normal limits of pre-DXM DS = 2.5 Mg/ml, T = 60 ng/dL While upper normal Limitsof post-DXM DS = 0.5 rg/ml, T = 25 ng/dl. bMean f S.E. ‘P < 0.0 1 when compared with the pm-DXM value. dP < 0.01 when compared with the normal follicular phase levels (DHAS: 1.61 f 0.12 &ml, T: 42 + 2.4 ngldl, u = 14). ZntJ Gynaecol Obstet 20
f 7.6 f 1.5’ f 3.6’ f 1.2c &4.oc
DXM suppression teest in hirsu tism
Results Fifty-one out of 55 patients who had gone through the DXM suppression test showed an adequate adrenal suppression (i.e. post-DXM DS less than 0.5 pg/ml). The data of the four patients with incomplete adrenal suppression (i.e. post-DXM DS higher than 0.5 E.cg/ml) were deleted. Among these fifty-one patients, 17 (33%) patients showed androgen excess of ovarian source (ovarian group), 13 (26%) patients had adrenal source (adrenal group), while 18 (35%) patients had a mixed ovarian and adrenal source (mixed group) of androgen Table II.
3I
excess. Three patients (6%) out of 5 1 patients failed to identify the source (none group) of androgen excess (Table I). Therefore, the source of androgen excess was identified by this DXM suppression test in 94% (48 out of 51) of hirsute women. Elevated pre-DXM DS mean level was noted in the hirsute women and this was mainly attributed by the elevated level in the adrenal and mixed group. In the group of ovarian androgen excess, the DS level was within normal range. The preDXM T level was elevated in all groups except with unidentifiable source the patients (none group) of androgen excess. Post-DXM T
Distribution of the clinical findings in hirsute females,
Clinical fmding
All patients
Source of androgen excess ovarian
Adrenal
Mixed __-_!-
None
51a(100)b (lOOjC
17(100) (33)
13(100) (26)
18(100) (35)
3(100) (6)
9(18) (100) 28(55) (100) 14(27) (100)
3(18) (33) lO(59) (36) 4(23) (2%
4(31) (45) 4(31) (14) 5(38) (35)
2(11) (22) 12(67) (43) 4(22) (29)
0) (0) 2(67) (7) l(33) (7)
29(57) (100) 15(29) (100) 7(14) (100)
8(47) (27) 3(18) (20) 6(35) (86)
8(61) (27) 4(31) (27) l(8) (14)
12(67) (42) 6(33) (40) O(O) (0)
I(331 (4) 2(67) (13) O(O) (0)
Acne
16(31) (100)
2(12) (13)
6(46) (37)
7(39) (44)
I(331 (6)
Male escatcheon
16(31) (100)
7(41) (44)
3(23) (19)
6(33) (37)
O(O) (0)
4(8) (100)
2(12) (50)
O(O) (0)
2(11) (50)
O(O) (0)
13(26) (100)
8(47) (62)
O(O) (0)
5(28) (38)
O(O) (0)
-..(All patients)
Menstrual pattern: Eumenorrhea Oligomenorrhea Amenorrhea
Hirsutism: Mild Moderate Severe
Clitoromegaly
Polycystic ovaries
aNumber of patient. b% in the vertical group. ‘% in the horizontal group. Int J Gynaecol Obstet 20
38
Wu
level was suppressed in all groups of patients except the ovarian group. In order to evaluate the relationship between the menstrual patterns and the clinical signs of androgen excess vs. the sources of androgen excess, the percent distribution of all the clinical signs were tabulated according to the source of androgen excess (Table II). Eighteen percent of total patients had eumenorrhea but oligomenorrhea was the most common menstrual pattern observed in hirsute women. Mild to moderate hirsutism was a common occurrence in these patients and only 31% of patients showed evidence in acne. Male escatcheon was observed in 16 patients (3 1%), whereas clitoromegaly was only observed in four patients (8%). Clinical diagnosis of polycystic ovarian disease was noted in 26% of the patients. When these clinical findings were distributed accordingly to their source of androgen excess, eumenorrhea was common in the adrenal group and oligomenorrhea-amenorrhea was more frequently observed in the ovarian or mixed groups. Severe hirsutism was quite common in the ovarian group, whereas mild hirsutism is more frequently seen in the adrenal and the mixed groups. Acne was less Table III.
common in the ovarian group, and male escatcheon associated less frequently with the adrenal group. Polycystic ovary was not seen in the adrenal group, yet common in the ovarian and mixed groups. The baseline DS and T level in this group of hirsute women was reevaluated retrospectively to see the frequency of abnormalities in the different groups of androgen excess (Table III). Nomzal DS and T Zevel was noted in 11 patients (22%) out of total 5 1 hirsute women and more than half (55%) of them belonged to the ovarian group. Only 7 patients had elevated DS and normal T level and most of them belonged to the mixed group and none in the ovarian group. Approximately one third of the patients showed normal DS and elevated T. Most of them (64%) were in the ovarian group. Approximately another one third (3 1%) of patients had both DS and T elevated. This latter combination was more frequently observed in the mixed group and none in the ovarian group. Even an elevated DS and/or T was considered, only 78% of hirsute women showed an abnormality, although, 100% of mixed group and 85% of adrenal group had abnormalities in this category.
Distribution of the baseline DHA-sulfate (DS) and testosterone (T) levels in hirsute females.
source of androgen excess
Androgen level DS: -a T: -
DS: tb T: -
DS: T: t
DS: t T: t
DS: f and/or T: t
llc(lOO)d we 6 (55) (35) 2(18) (15) O(0) (0) 3(27) (100)
7(100) (14) O(0) (0) X29) (15) 5(71) (28) O(0) (0)
17(100) (33) 1 l(64) (65) 4(24) (31) 202) (11) O(0) (0)
16(100) (31) O(0) (0) 5(31) (39) 1 l(69) (61) O(0) (0)
40(100) (78) 1 l(28) (65) 1 l(28) (85) 18(44) (100) O(0) (0)
-All patients n=Sl Ovarian n-17 Adrenal n=13 Mixed n=18 None n=3 a
-,
normal.
bt; elevated. %mber of patient. + Din the vertical group. %/oin the horizontal group. In? J Gynaecol Obstet 20
DXMsuppression
Discussion Dexamethasone suppression tests have been utilized to search for the source of androgen excess in hirsute women for the last couple of decades. Its value has been questioned or affiied [4,10]. One of the major problems encountered in previous DXM suppression tests was the inconsistancy of urinary 17ketosteroids in reflecting the androgen state. 17-Ketosteroids are androgen metabolites derived from adrenal as well as ovarian androgen; thus it does not exclusively represent the androgen activity of either gland [ 10,141. In addition, the dosage of DXM and the duration of medication were arbitrarily chosen without evaluating its adequacy. Thus interpretation of results can be clouded by data obtained under incomplete adrenal suppression. More than 90% of circulating DS is originated from the adrenal [ 1,3]. Its peripheral concentration is stable and its metabolic half-life is much longer [ 13,191, therefore, DS fluctuates less than cortisol under acute stress and it reflects better the adrenal androgen secretory activity. Although DS is ACTH dependent but its response to ACTH is not as rapid nor sensitive as corticosteroids [ 191. Perhaps adrenal androgen stimulating hormone (AASH), a possible new hormone has not been identified, or prolactin may stimulate the adrenal androgen secretion [ 7,191. During the DXM administration, cortisol level was suppressed to the baseline level within 2 or 3 days, yet DS would not reach a basal level during this period of time. Instead it takes approximately 2 weeks for DS to reach the baseline level [ 1,101. Therefore, it is advisable to administer DXM for 2 weeks in order to have a completely suppressed adrenal androgen secretion. Incomplete adrenal suppression occurring in a 2-day suppression test may complicate the interpretation of the result. Even in a 1 week DXM suppression, there were only 54% of patients who showed an adequate adrenal suppression [4]. In addition, the dosage of DXM should be varied
test in hirsutism
39
according to the body weight as suggested by Abraham [4]. This is especially true in hirsute women as they tend to have excessive weight gain. The sources of excess androgen in hirsute women was evaluated by measuring plasma concentration of androgens in the adrenal and ovarian vein blood [ 11,121. Its result suggested that in more than 90% of hirsute women their androgen excess was originated from the ovary. This elegant study, however, was questioned for its result because of the subsequent evidence disputed their theoretical assumption of a parallel secretory activity between adrenal androgen and cortisol [2]. Furthermore, the episodic secretory pattern of adrenal steroids [ 101 makes it difficult to interpretate the long-term adrenal function on the basis of data in venous blood samples obtained by catheterization. It’s finding that DXM suppresses ovarian androgen secretion was also questioned, because the androgen level in long-term DXM treated women was not different from those levels in the adrenalectomized women [ 1,3]. Therefore, DXM can be properly used to suppress adrenal androgen secretion in order to define the source of androgen excess in hirsute women. Adrenal androgen excess state may lead to a disrupted ovarian folliculogenesis and subsequently increases ovarian androgen secretion. Therefore, it is quite conceivable that initial adrenal androgen hyper-secretion may lead into a mixed ovarian and adrenal hypersecretion. The possibility of opposite phenomenon, i.e. initial ovarian androgen excess induces subsequent adrenal androgen excess, is not clear at the present time. However, in those patients with mixed androgen excess, it is possible to reverse ovarian androgen excess state, especially when patient becomes ovulatory. Because of this possibility of changing the source of androgen excess, it should be flexible in treating hirsute women and reevaluation may be necessary from time to time. A partial 1 I- and 21-hydroxylase deficiency was reported in a group of hirsute women with adrenal androgen excess [ 161. Iat J Gynaecol Obstet 20
40
wu
Table N.
Sources of androgen excess in hirsute women.
References
Patient No.
Kirschner and Jacobs [ 11 ] Givens et al. [9] Kirschner et al. [ 121 Toaff et al. [18] Abraham et al. [5] Wu (present study) TOTAL
Ovarian
Adrenal
Mixed
13 20 44 35 86 48
9/l 3(69%) 8/20(40%) 42/44(95%) 2/35(6%) 16/86(19%) 17/48(35%)
o/ 13(0%) 2/20(10%) 2/44(5%) 25/35(71%) 44/86(5 1%) 13(48(27%)
4/13(31%) 10/20(50%) O/44(0%) 8/35(23%) 26/86(30%) 18/48(38%)
246
94/246(38%)
86/246(35%)
66/246(27%)
Thus hirsute women should be carefully evaluated to rule out the mild or “attenuated” congenital adrenal hyperplasia [ 171. Oligomenorrhea and/or amenorrhea were the most common menstrual patterns in hirsute women. Eumenorrhea was more common in women with adrenal androgen excess. Perhaps the latter group of patient may still preserve ovulatory function, inspite of androgen excess on some occasions. Thus a mild adrenal hyper-androgen secretion may not disrupt ovulation. The patients with ovarian androgen excess tend to develop severe hirsutism. This might be related to the elevated T which was more frequently observed in this group of patients. Acne was seen in 31% of patients and most of them suggested adrenal origin of androgen excess. Thus, adrenal androgen is probably related to acne formation in hirsute women. Male escatcheon was evenly observed among the three groups and clitoromegaly was only observed in four patients thus there were too few patients to evaluate. Polycystic ovary was noted in 26% of patients. This low incidence observed in this study might be due to some patients who did not have laparoscopic examination or LH determination. It is possible that a higher incidence of polycystic ovary may occur in hirsute women. Source of androgen excess in hirsutism is among the probably evenly distributed adrenal, ovarian, and mixed groups (Table IV). Since the mixed group can be included in either the adrenal or the ovarian origin, it is Int J Gynaecol
Obstet 20
Source of androgen excess
quite acceptable to consider either the ovary or the adrenal as the major source of androgen excess in hirsutism. This study showed that measurement of DS and T before and after the 2-week DXM suppression can identify the source of androgen excess in 94% of hirsute women. This can be a simple and practical test in evaluation of hirsute women. References Abraham GE: Ovarian and adrenal contribution to peripheral androgens during the menstrual cycle. J CJin Endocrinol Metab 39: 340,1974. Abraham GE: Ovarian and adrenal contribution to peripheral steroids during the menstrual cycle in two hirsute women. Obstet Gynecol. 46: 29,1975. Abraham GE, Chakmakjian ZH: Serum steroid levels during the menstrual cycle in a bilaterally adrenalectomized woman. J Clin Endocrinol Metab 27: 581, 1973. Abraham GE, Maroulis GB, Buster JE et al: Effect of dexamethasone on serum cortisol and androgen levels in hirsute patients. Obstet Gynecol47: 395,1976. Abraham GE, Maroulis GB, Boyers SP et al: Dexamethasone suppression test in the management of hyperandrogenized patients. Obstet Gynecol5 7: 158,198 1. Bardin CW, Lipsett MB: Testosterone and androstenedione blood production rates in normal women and women with idiopathic hirsutism or polycystic ovaries. J Clin Endocrinol Metab 46: 891,1967. Bassi F, Giusti G, Borsi L et al: Plasma androgens in women with hyperprolactinaemic amenorrhea. Clin Endrocrinol6: 5,1977. Buster JE, Abraham GE: Radioimmunoassay for dehydroepiandrosterone sulfate. Anal Lett 5: 543: 1972. Givens JR, Andersen RN, Ragland JB et al: Andrenal function in hirsutism I. Diurnal change and response of plasma androstenedione, testosterone, 17-hydroxyprogesterone, cortisol, LH and FSH to dexamethasone and
DXM suppression test in him tism
10
11
12
13
14
15
16
yZ unit of ACTH. J Clin Endocrinol Metab 40: 988, 1975. Judd HL, McPherson RA, Rakoff JS et al: Correlation of the effects of dexamethasone administration on urinary 17ketosteroids and serum androgen levels in patients with hirsutism. Am J Obstet Gynecol 1.28: 408,1977. Kirschner MA, Jacobs JB: Combined ovarian and adrenal vein catheterization to determine the site(s) of androgen overproduction in hirsute women. J Clin Endocrinol Metab 33: 199,197l. Kirschner MA, Zucker IR, Jespersen, D: Idiopathic hirsutism - an ovarian abnormality. N Engl J Med 294: 637,1976. KorthSchutz S, Levine LS, New MI: Dehydroepiandrosterone sulfate (DS) levels, a rapid test for abnormal adrenal androgen secretion. J Clin Endocrinol Metab 42: 1005,1976. Lobo RA, Paul W, Goebelsmann U: Clinical use of plasma DHEA-sulfate: correlations with total and fractionated urinary 17-ketosteroids. The pacific mast Fertility Society Meeting 1979. Abstract p. 27. Maroulis GB, Manlimos FS, Abraham GE: Comparison between urinary 17ketosteroids and serum androgens in hirsute patients. Obstet Gynecol49: 454,1977. Newmark S, Dluhy RG, Williams GH et al: Partial lland 21-hydroxylase deficiencies in hirsute women. Am J Obstet Gynecoll27: 594, 1977.
41
17 Rosenwaks Z, Lee PA, Jones GS et al: An attenuated form of congenital adrenal hyperplasia. J Clin EndocrinolMetab49: 335,1979. 18 Toaff, R, Toaff ME, Gould S et al: Role of androgenie hyperactivity in anovulation. Fertil Steril 29: 407,1978. 19 Vaitukaitis JL, Cale SL, Melby JC: Role of ACTH in the secretion of free dehydroepiandrosterone and its sulfate ester in man. J Clin Endocrinol Metab 29: 1443,1969. 20 Wu CH, Blasco L, Flickinger GL et al: Ovarian function in the preovulatory rabbit. Biol Reprod 17: 304, 1977. 21 Yen SSC, Jaffe RB: Reproductive Endocrinology, physiology, pathophysiology and clinical management, p. 301. W.B. Saunders Company, 1978.
Address for reprints:
Chung H. Wu, M.D. Dept of Obstetrics and Gynecology Jefferson Medical College Room 300 1025 Walnut Street Philadelphia, PA 19107 USA
Znt J Gynaecol Obstet 20