Adrenal and ovarian steroid hormone responses to gonadotropin-releasing hormone agonist treatment in polycystic ovary syndrome Frank Gonzalez, MD, Denise A. Hatala, BS, and Leon Speroff, MD Cleveland, Ohio It has been postulated that in polycystic ovary syndrome ovarian steroids can influence adrenal steroidogenesis. To test this hypothesis, basal and dexamethasone-suppressed-corticotropin-stimulated steroid hormone responses were compared among three groups of women before, during, and after gonadotropin-releasing hormone agonist treatment for 3 months. The groups were characterized as follows: (1) women with polycystic ovary syndrome with high dehydroepiandrosterone sulfate levels (>400 f1g/dl), (2) women with polycystic ovary syndrome with normal dehydroepiandrosterone sulfate levels «300 f1g/dl), and (3) normal ovulatory women. In response to gonadotropin-releasing hormone agonist, basal serum luteinizing hormone, follicle-stimulating hormone, estradiol, estrone, 17-hydroxyprogesterone, androstenedione, and testosterone in all three groups were suppressed to similar levels. Basal serum dehydroepiandrosterone sulfate levels in the group with high levels declined, but they did not reach the normal, unaltered concentrations in the other two groups. Two subjects with polycystic ovary syndrome in this group with high levels, who showed the greatest declines in basal serum dehydroepiandrosterone sulfate levels (34%, 40%), also had evidence of 3[3-hydroxysteroid dehydrogenase deficiency before treatment, which was resolved by the end of treatment. In both groups with polycystic ovary syndrome, the increase in maximum incremental rise of dehydroepiandrosterone and dehydroepiandrosterone sulfate levels in response to a pharmacologic dose of corticotropin from a dexamethasone-suppressed baseline (adrenal androgen capacity) remained unaltered during gonadotropin-releasing hormone agonist administration. We conclude that ovarian steroids may promote excessive adrenal androgen secretion in women with polycystic ovary syndrome, may induce 3[3-hydroxysteroid dehydrogenase deficiency as a mechanism for adrenal involvement in some women with polycystic ovary syndrome, and do not influence adrenal androgen capacity. (AM J OSSTET GVNECOL 1991 ;165:535-45.)
Key words: Polycystic ovary syndrome, gonadotropin-releasing hormone agonist, dehydroepiandrosterone sulfate, 313-hydroxysteroid dehydrogenase deficiency
The mechanism of excess adrenal androgen secretion in polycystic ovary syndrome remains controversial. It has been postulated that ovarian steroids influence adrenal steroidogenesis by inhibition of adrenal 313-hydroxysteroid dehydrogenase activity or by causing adrenal androgen hyperresponsiveness to corticotropin. This hypothesis may account for the elevations in serum levels of dehydroepiandrosterone sulfate (DHA-S) observed in approximately half of women with polycystic ovary syndrome. 1,2 Administration of a
From the Department of Reproductive Bi%gy, School of Medicine, Case Western Reserve University, Supported in part by General Clinical Research Center grant MOl RR00080 from the Division of Research Resources, National Institutes of Health, to University Hospitals, Cleveland, Ohio, Presented at the Thirty-seventh Annual Meeting of the Society for Gynecologic Investigation, St, Louis, Missouri, March 21-24, 1990, Received for publication November 7, 1990; revised February 27, 1991; accepted March 15, 1991, Reprint requests: Frank Gonzalez, MD, Children's Hospital of Buffalo, Division of Reproductive Endocrinology, 219 Bryant St., Buffalo, NY 14222. 6/1129544
long-acting gonadotropin-releasing hormone (GnRH) agonist has been reported to selectively suppress pituitary gonadotropins and secretion of ovarian steroids in women with polycystic ovary syndrome.'" However, the effect of administering GnRH agonist to women with polycystic ovary syndrome with high DHA-S levels (>400 ILg / dl) has never been examined. Thus this study was undertaken to determine if women with polycystic ovary syndrome and high DHA-S levels have a different adrenal response than do women with polycystic ovary syndrome and normal DHA-S levels to GnRH agonist administration for 3 months. Basal and dexamethasone - suppressed -corticotropin -stimulated hormone profiles were evaluated before, during, and after treatment, and the results were compared with those in a group of normal ovulatory women who were similarly evaluated for the same period of time.
Material and methods Subjects. Ten women with polycystic ovary syndrome were selected for study because they showed elevation in at least one serum androgen level (DHA-S, >340
535
536
Gonzalez, Hatala, and Speroff
September 1991 Am J Obstet Gynecol
Table I. Baseline clinical data in 10 women with polycystic ovary syndrome Ovarian appearance Age (yr)
Ideal body weight (%)
Menses (cycles Iyr)
Hirsutism
10
18 25 22 19 31
142 182 173 126 176
3 6 6 2 3
+ + + + +
Normal DHA-S It 3 5 6 9
23 33 35 35 27
173 168 139 212 120
7 5 3 1 6
+ + + + +
Subject No.
High DHA-S 2 4 7 8
Enlarged*
+
1
Polycystid
+ +
+ + + + + +
+ + + +
+
+, Positive; -, negative. *Both ovaries >4.5 em in length. t Both ovaries with more than 5 follicles < 10 mm in diameter. :j:Aeanthosis nigrieans. f.Lg/dl; total testosterone, >0.7 ng/ml, or free testosterone, >0.8 ng/ml) and exhibited the classic features of oligomenorrhea and hirsutism. All had normal prolactin levels «20 f.Lg/L), were euthyroid (thyroidstimulating hormone, 120% of ideal body weight), and one had acanthosis nigricans. Although all subjects with polycystic ovary syndrome underwent vaginal ultrasonography for ovarian characterization before the study, enlarged or polycystic ovaries were not required for the diagnosis of polycystic ovary syndrome. In retrospect, all exhibited low-dose dexamethasone suppression of cortisol to 2 SD above the mean of ovulatory women were considered to have 313-hydroxysteroid dehydrogenase deficiency. Ratios of the maximum absolute values of 17 -hydroxypregnenolone to DHA or 17 -hydroxyprogesterone to androstenedione that were >2 SD below the mean of normal ovulatory women were required to fulfill the criteria for 17,20-lyase hyperactivity. Cortisol output was defined as the maximum absolute value of cortisol secreted. The maximum incremental rise in DHA and DHA-S secretion from a standardized dexamethasone-suppressed baseline (mean of values at - i 5 and 0 minutes) was calculated to represent adrenal androgen capacity. Alterations in these parameters during GnRH agonist administration were determined from adrenal dynamic testing performed after 1 month and 3 months of treatment. Statistical evaluation was performed with two-way analysis of variance, followed by Student paired or group t tests, where appropriate. One-tailed significance was considered sufficient when paired comparisons were assessed for the basal serum levels of LH, FSH, estradiol, estrone, 17-hydroxyprogesterone, androstenedione, and testosterone because only declines in these hormones were anticipated in response to GnRH agonist, as demonstrated previously.3. 4. 7. 8 However, two-tailed significance was used for all other comparisons. The Bonferroni test was chosen to control for multiple comparisons when differences across groups during treatment were assessed. 9 In this instance a value of 0.0 17 was used to detect a significant difference at the level of 0.05. Linear regression was used to examine correlations in the responses to GnRH agonist among the basal serum levels of estradiol, estrone, and DHA-S.
538
Gonzalez, Halala, and Speroff
Se pLelll her 1991 A ll! J OhsLel
I I I
~J
3
Months GnRH-a
Fig. 1. Basal serum LH , FSH, estradiol, and estron e concemrations (mean ± SE) in women with polycystic ovary syndrome widl high and normal DHA-S levels and in ovulatory women before and durin g GnRH agonist (GnRH-a) treatment. FaClors used to convert data into System International d 'U nites (51 units) are 3.67 1 and 3.699 for estradi ol and estrone, respecti vely.
Results
Basal hormonal response Baseline hormone concentrations. The baseline hormonal characteristics of the tht'ee treatment groups are shown in Table II. The mean serum concentrations of a ndrostenedione, testosterone, DHA, and DHA-S just before treatment were significantly higher (P < 0.(5) in the high DHA-S group than in ovulatory women. Baseline hormonal differences between the polycystic ovary syndrome groups we re limited to those in mean serum DHA and DHA-S levels , which were significantly higher (p < 0.05) in the high DHA-S group than in the normal DHA-S group. Both polycystic ovary syndrome groups demonstrated elevations in serum androgens and elevated LH to FSH ratios of 2.0 and 1.7 in the high DHA-S group and the normal DHA-S group, respectively. A reversal of the normal estradiol to estrone ratio also was evident in the normal DHA-S group. Gonadotropins and estrogens. As shown in Fig. 1, basal gonadotropin levels (mean j: SE) before and during GnRH agonist administration in the high and normal DHA-S groups and ovulatory women are compared
with each other. Administration of GnRH agonist for I month resulted in marked and significant reductions (p < 0.(5) of mean serum LH and FSH levels in both groups with polycystic ovary syndrome. These reductiolls were maintained throughout the entire treatment pe riod. In ovulatory women, gonadotropin concentrations exhibited only modest declines. However, respective serum levels of LH and FSH were similar in all three groups throughout the course of GnRH agonist therapy. The serum estradiol and es trone responses to GnRH agonist after 1 month of treatment (Fig. I) also were characterized by marked and signi ncant reductions (p < 0.05) in all three groups. Respective serum levels of estrad iol and estrone remained low and were simila r in all three groups throughout the treatment period . In the case of serum estradiol, levels during treatment approached the limits of sensitivity (10 pg/ ml or 36 pmollL) for the assay. Progestirt.l. The basal serum progestin levels (mean ± SE) before and during GnRH agonist treatment are shown in Fig. 2. Serum 17-hydroxypregnen-
Adrenal response to GnRH in polycystic ovary syndrome
Volume 165 Number 3
7
I --
~--------------------~l
.J
E
't:n
c:
6
High DHA'S
----
Normal
DHA~S
...r::, ••• Ovulatory
0 0
c Q) c
....
0..
E c
5
Q)
c
e
4
Q)
1ii Q)
3
t:n
o...
t:n
Q)
.J
"t:n \.5
Q)
c
I
539
2
0.. 0.5
J:
J:
o
0
I'-
0
L-_~
_ _ _ _ __- L_ _ _ _ _ _J -_ _
o
~
I
2
~_~
___ _
t: o
3
1
2
Months GnRH-a
Months GnRH-o
I>
14
5
.J 1.2
r-------------.------~
.J
E
'-
E
t:n
c
"-
CI
4
C
Q)
c
.2
"Q
Q)
c
3
.! 06 '"o 04 1ii
Q)
C
.... Q)
'"...
2
0
~
"Q
C
Key words: Polycystic ovary syndrome, gonadotropin-releasing hormone agonist, dehydroepiandrosterone sulfate, 313-hydroxysteroid dehydrogenase deficiency
The mechanism of excess adrenal androgen secretion in polycystic ovary syndrome remains controversial. It has been postulated that ovarian steroids influence adrenal steroidogenesis by inhibition of adrenal 313-hydroxysteroid dehydrogenase activity or by causing adrenal androgen hyperresponsiveness to corticotropin. This hypothesis may account for the elevations in serum levels of dehydroepiandrosterone sulfate (DHA-S) observed in approximately half of women with polycystic ovary syndrome. 1,2 Administration of a
From the Department of Reproductive Bi%gy, School of Medicine, Case Western Reserve University, Supported in part by General Clinical Research Center grant MOl RR00080 from the Division of Research Resources, National Institutes of Health, to University Hospitals, Cleveland, Ohio, Presented at the Thirty-seventh Annual Meeting of the Society for Gynecologic Investigation, St, Louis, Missouri, March 21-24, 1990, Received for publication November 7, 1990; revised February 27, 1991; accepted March 15, 1991, Reprint requests: Frank Gonzalez, MD, Children's Hospital of Buffalo, Division of Reproductive Endocrinology, 219 Bryant St., Buffalo, NY 14222. 6/1129544
long-acting gonadotropin-releasing hormone (GnRH) agonist has been reported to selectively suppress pituitary gonadotropins and secretion of ovarian steroids in women with polycystic ovary syndrome.'" However, the effect of administering GnRH agonist to women with polycystic ovary syndrome with high DHA-S levels (>400 ILg / dl) has never been examined. Thus this study was undertaken to determine if women with polycystic ovary syndrome and high DHA-S levels have a different adrenal response than do women with polycystic ovary syndrome and normal DHA-S levels to GnRH agonist administration for 3 months. Basal and dexamethasone - suppressed -corticotropin -stimulated hormone profiles were evaluated before, during, and after treatment, and the results were compared with those in a group of normal ovulatory women who were similarly evaluated for the same period of time.
Material and methods Subjects. Ten women with polycystic ovary syndrome were selected for study because they showed elevation in at least one serum androgen level (DHA-S, >340
535
536
Gonzalez, Hatala, and Speroff
September 1991 Am J Obstet Gynecol
Table I. Baseline clinical data in 10 women with polycystic ovary syndrome Ovarian appearance Age (yr)
Ideal body weight (%)
Menses (cycles Iyr)
Hirsutism
10
18 25 22 19 31
142 182 173 126 176
3 6 6 2 3
+ + + + +
Normal DHA-S It 3 5 6 9
23 33 35 35 27
173 168 139 212 120
7 5 3 1 6
+ + + + +
Subject No.
High DHA-S 2 4 7 8
Enlarged*
+
1
Polycystid
+ +
+ + + + + +
+ + + +
+
+, Positive; -, negative. *Both ovaries >4.5 em in length. t Both ovaries with more than 5 follicles < 10 mm in diameter. :j:Aeanthosis nigrieans. f.Lg/dl; total testosterone, >0.7 ng/ml, or free testosterone, >0.8 ng/ml) and exhibited the classic features of oligomenorrhea and hirsutism. All had normal prolactin levels «20 f.Lg/L), were euthyroid (thyroidstimulating hormone, 120% of ideal body weight), and one had acanthosis nigricans. Although all subjects with polycystic ovary syndrome underwent vaginal ultrasonography for ovarian characterization before the study, enlarged or polycystic ovaries were not required for the diagnosis of polycystic ovary syndrome. In retrospect, all exhibited low-dose dexamethasone suppression of cortisol to 2 SD above the mean of ovulatory women were considered to have 313-hydroxysteroid dehydrogenase deficiency. Ratios of the maximum absolute values of 17 -hydroxypregnenolone to DHA or 17 -hydroxyprogesterone to androstenedione that were >2 SD below the mean of normal ovulatory women were required to fulfill the criteria for 17,20-lyase hyperactivity. Cortisol output was defined as the maximum absolute value of cortisol secreted. The maximum incremental rise in DHA and DHA-S secretion from a standardized dexamethasone-suppressed baseline (mean of values at - i 5 and 0 minutes) was calculated to represent adrenal androgen capacity. Alterations in these parameters during GnRH agonist administration were determined from adrenal dynamic testing performed after 1 month and 3 months of treatment. Statistical evaluation was performed with two-way analysis of variance, followed by Student paired or group t tests, where appropriate. One-tailed significance was considered sufficient when paired comparisons were assessed for the basal serum levels of LH, FSH, estradiol, estrone, 17-hydroxyprogesterone, androstenedione, and testosterone because only declines in these hormones were anticipated in response to GnRH agonist, as demonstrated previously.3. 4. 7. 8 However, two-tailed significance was used for all other comparisons. The Bonferroni test was chosen to control for multiple comparisons when differences across groups during treatment were assessed. 9 In this instance a value of 0.0 17 was used to detect a significant difference at the level of 0.05. Linear regression was used to examine correlations in the responses to GnRH agonist among the basal serum levels of estradiol, estrone, and DHA-S.
538
Gonzalez, Halala, and Speroff
Se pLelll her 1991 A ll! J OhsLel
I I I
~J
3
Months GnRH-a
Fig. 1. Basal serum LH , FSH, estradiol, and estron e concemrations (mean ± SE) in women with polycystic ovary syndrome widl high and normal DHA-S levels and in ovulatory women before and durin g GnRH agonist (GnRH-a) treatment. FaClors used to convert data into System International d 'U nites (51 units) are 3.67 1 and 3.699 for estradi ol and estrone, respecti vely.
Results
Basal hormonal response Baseline hormone concentrations. The baseline hormonal characteristics of the tht'ee treatment groups are shown in Table II. The mean serum concentrations of a ndrostenedione, testosterone, DHA, and DHA-S just before treatment were significantly higher (P < 0.(5) in the high DHA-S group than in ovulatory women. Baseline hormonal differences between the polycystic ovary syndrome groups we re limited to those in mean serum DHA and DHA-S levels , which were significantly higher (p < 0.05) in the high DHA-S group than in the normal DHA-S group. Both polycystic ovary syndrome groups demonstrated elevations in serum androgens and elevated LH to FSH ratios of 2.0 and 1.7 in the high DHA-S group and the normal DHA-S group, respectively. A reversal of the normal estradiol to estrone ratio also was evident in the normal DHA-S group. Gonadotropins and estrogens. As shown in Fig. 1, basal gonadotropin levels (mean j: SE) before and during GnRH agonist administration in the high and normal DHA-S groups and ovulatory women are compared
with each other. Administration of GnRH agonist for I month resulted in marked and significant reductions (p < 0.(5) of mean serum LH and FSH levels in both groups with polycystic ovary syndrome. These reductiolls were maintained throughout the entire treatment pe riod. In ovulatory women, gonadotropin concentrations exhibited only modest declines. However, respective serum levels of LH and FSH were similar in all three groups throughout the course of GnRH agonist therapy. The serum estradiol and es trone responses to GnRH agonist after 1 month of treatment (Fig. I) also were characterized by marked and signi ncant reductions (p < 0.05) in all three groups. Respective serum levels of estrad iol and estrone remained low and were simila r in all three groups throughout the treatment period . In the case of serum estradiol, levels during treatment approached the limits of sensitivity (10 pg/ ml or 36 pmollL) for the assay. Progestirt.l. The basal serum progestin levels (mean ± SE) before and during GnRH agonist treatment are shown in Fig. 2. Serum 17-hydroxypregnen-
Adrenal response to GnRH in polycystic ovary syndrome
Volume 165 Number 3
7
I --
~--------------------~l
.J
E
't:n
c:
6
High DHA'S
----
Normal
DHA~S
...r::, ••• Ovulatory
0 0
c Q) c
....
0..
E c
5
Q)
c
e
4
Q)
1ii Q)
3
t:n
o...
t:n
Q)
.J
"t:n \.5
Q)
c
I
539
2
0.. 0.5
J:
J:
o
0
I'-
0
L-_~
_ _ _ _ __- L_ _ _ _ _ _J -_ _
o
~
I
2
~_~
___ _
t: o
3
1
2
Months GnRH-a
Months GnRH-o
I>
14
5
.J 1.2
r-------------.------~
.J
E
'-
E
t:n
c
"-
CI
4
C
Q)
c
.2
"Q
Q)
c
3
.! 06 '"o 04 1ii
Q)
C
.... Q)
'"...
2
0
~
"Q
C
