J. Steroid Biochem. Molec. Biol. Vol. 42, No. 2, pp. 211-221, 1992
0960-0760/92 $5.00 + 0.00 Copyright © 1992 Pergamon Press pie
Pr~dled in Great Britain. All rights reserved
PLASMA LEVELS OF C19 STEROID GLUCURONIDES IN PRE-MENOPAUSAL WOMEN WITH NON-CLASSICAL CONGENITAL A D R E N A L HYPERPLASIA C. B. WHORWOOD,* H. UESHIBAand P. DEL BALZO Division of Pediatric Endocrinology, The New York Hospital-Cornell Medical Center, New York, NY 10021, U.S.A.
(Received 3 July 1991)
Summary--Recent reports have thrown doubt on the role of measurements of plasma 5~t-androstane-3~t,17fl-diol glucuronide (3~t-diolG) as a marker of peripheral androgen metabolism in women with polycystic ovarian syndrome and idiopathic hirsutism. It has been suggested that a plasma profile of C~9 steroid glucuronides may be more informative. While preliminary data indicates that both 3~t-diolG and androsterone G (ADTG) may arise from adrenal steroid precursors, there have been no reports of C~9 steroid glucuronides in women with non-classical, or late-onset congenital adrenal hyperplasia (NC-CAH), who constitute a significant proportion of the hirsute female population. We therefore measured plasma levels of 3~t-diolG, ADTG and dihydrotestosterone G (DHTG) before and following a standard Cortrosyn test in 15 symptomatic and 3 asymptomatic NC-CAH patients, 5 heterozygote carriers for 21-hydroxylase deficiency (NCHETS) and 18 normal women. The effects of chronic glucocorticoid (GCR) therapy (> 3 months) on the C~9 steroid glueuronide profile in the symptomatic patients was also investigated. Baseline plasma levels of all 3 glucuronides were significantly (P < 0.001) higher in symptomatic patients compared with either normals or NCHETS. However, the order of discrimination was A D T G > 3~t-diolG > DHTG. There were no significant differences between steriod glucuronide levels for NCHET and normal women and the C~9 steroid glucuronide concentrations for the asymptomatic NC-CAH patients were > 2 SD above the normal means. Moderate clinical improvement was observed in all patients receiving oral GCR therapy and was accompanied by approx. 80% suppression of the plasma levels of all 3 Ct9 steroid glucuronides. This contrasts with a mean suppression of androstenedione of only 50%. However, plasma levels of the C~9 steroid glucuronides were not significantly increased in response to a short ACTH stimulation test. This may be explained by the fact that the androgen glucuronides are thought to be peripherally formed metabolites derived from unconjugated glandular secreted androgen precursors and thus their synthesis at 60 rain following adrenal stimulation may lag substantially behind that of their respective precursors. There were significant linear correlations between the levels of all 3 glucuronides, but neither correlated with Ferriman--Gallway scores, body mass index or 17-hydroxyprogesterone levels. We therefore conclude that (1) plasma levels of 3~-diolG, ADTG and DHTG are neither related to hirsutism as previously suggested, nor body weight (in normal weight individuals) and since they are elevated in asymptomatic NC-CAH patients they may therefore only reflect adrenal contribution to circulating androgens and not peripheral androgen action per se; (2) a plasma profile of Cj9 steroid glucuronides may serve as an excellent index of the effectivenesss of treatment in hyperandrogenic disorders of primarily adrenal etiology; (3) the heterozygosity of NC-CAH does not appear to be expressed in terms of a gluco-conjugated steroid profile intermediate between that of homozygous unaffected and affected individuals; and (4) there may be a common pathway of C~9 steriod glucuronide metabolism, the precursors of which are likely to be of exclusively adrenal origin in non-obese women.
INTRODUCTION
It is well k n o w n that a relatively mild defect o f adrenal steroidogenic enzyme activity, i.e. 21*To whom correspondence should be addressed at: Department of Medicine, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, England.
hydroxylase, l l]~-hydroxylase, or 3~-hydroxysteroid dehydrogenase which manifests during either peri- or post-pubertal life, m a y constitute a significant cause o f hyperandrogenism in w o m e n [ I - 5 ] . Non-classical (NC) congenital adrenal hyperplasia ( N C - C A H ) due to mild 21-hydroxylase deficiency (21-OH) results in
211
212
C.B. WttORWOODet al.
markedly increased plasma androgen levels as a consequence of the accumulation of the C21 adrenal steroids above the 21-hydroxylase step in the metabolic pathways leading to cortisol, and is the most common autosomally recessive inherited disorder known to man [3, 5]. This allelic variant of classical CAH [1, 2] has the specific HLA associations of NC-21OH, B14 DR1, which are distinct from those frequently linked to the classical condition [6]. The clinical presentation of NC-21OH is extremely heterogenous. Symptoms may manifest at any age and include premature pubarche, accelerated growth velocity, advanced bone age and ultimate short stature in pre-adolescent individuals; and hirsutism, oligomenorrhea, acne, alopecia and infertility in adult women [2, 7, 8]. Furthermore, there is a small proportion of women whose endocrinological profile is consistent with NC-21OH and yet they are asymptomatic, often having been diagnosed as a consequence of both genetic and hormonal family studies of classical patients [9]. Clearly, those patients who manifest symptoms later in life represent a major proportion if not the entire population of asymptomatic juvenile NC-CAH subjects. The considerable qualitative and quantitative symptom variability between individuals presenting with hyperandrogenic disorders, such as NC-CAH and polycystic ovarian syndrome (PCOS), in the presence of similar circulating androgen levels, has led to the search for a suitable plasma marker of peripheral androgen metabolism and target organ sensitivity. The potent active androgen, dihydrotestosterone (DHT) and 5~-androstane-3~,17fl-diol (3~-diol) are both 5~-reduced metabolites of circulating testosterone (T), but plasma measurements fail to adequately distinguish the underlying metabolic abnormalities associated with the varied clinical presentation in hirsute women from the level of peripheral androgen action for non-hirsute women [10-12]. Plasma levels of free T may offer improved discrimination but this remains unreliable [13]. Several recent reports have suggested that plasma levels of 3~-diol glucuronide (3ct-diolG) and, more recently, androsterone glucuronide (ADTG) may reflect peripheral androgen metabolism in hirsute women with either PCOS or idiopathic hirsutism (IH)[10, 14-19]. However, Scanlon et al. [20] and others[21,22] were unable to find any significant differences between plasma 3~t-diolG levels in hirsute vs asymptomatic
women with PCOS or between normal and IH women [18, 22] thereby throwing some doubt on the significance of plasma 3~-diolG measurements. Plasma concentrations of ADTG, on the other hand, were found to be several fold those of 3~-diolG and significantly elevated both in symptomatic women with PCOS [19] and in IH patients [18] compared with non-hirsute women with PCOS. It has therefore been suggested that a C19 steroid glucuronide profile may be more informative with respect to an analysis of peripheral androgen action and furthermore may indicate the preferred metabolic pathways [18], which may be relevant in assessing appropriate specific treatment. Recent evidence suggests that both 3~tdiolG and ADTG may be derived almost exclusively from adrenal androgen precursors [18, 20, 23, 24]. We therefore decided to measure the levels of these steroid glucuronides and those of DHTG, a likely precursor of circulating 3~t-diolG, in blood obtained from (1) untreated symptomatic women with NC-21OH; (2) asymptomatic women with NC-21OH; (3) women diagnosed as being heterozygote carriers for NC-21OH; and (4) normal women. In addition, the effects of a short ACTH [1-24] test and long-term glucocorticoid therapy upon the plasma steroid glucuronide profile were also studied.
MATERIALS AND METHODS
Subjects
The diagnosis of NC-21OH is well established. Eighteen hyperandrogenic women were diagnosed as suffering from this disorder either through initial presentation at our clinic with symptoms of androgen excess (n = 15) or through genetic and hormonal family studies (n = 3). A Cortrosyn test was performed and a 24 h urine collection obtained. The diagnosis was made on the basis of a significantly elevated urinary pregnanetriol excretion and baseline and post-Cortrosyn 17-hydroxyprogesterone (17OHP) and androstenedione (A4) levels within the NC-21OH domain of a nomogram defined by our previous studies. Subsequent HLA typing and DNA analysis confirmed the diagnosis [25]. A Cortrosyn test was also performed on 18 normal women and 5 compound heterozygote carriers for NC-21OH (NCHETS), who were either the mother or sibs of affected individuals.
S t e r o i d g l u c u r o n i d e s in N C - C A H
All the women in the study were 14--45 years of age. There was no significant age difference between any of the groups. The FerrimanGallway (FG) score [26] for the normal women, asymptomatic NC-21OH and NCHETS was < 5, while that for the NC-21OH patients was 17 ___4. The body mass index [BMI: weight (kg)/height (mE)] was normal (
0960-0760/92 $5.00 + 0.00 Copyright © 1992 Pergamon Press pie
Pr~dled in Great Britain. All rights reserved
PLASMA LEVELS OF C19 STEROID GLUCURONIDES IN PRE-MENOPAUSAL WOMEN WITH NON-CLASSICAL CONGENITAL A D R E N A L HYPERPLASIA C. B. WHORWOOD,* H. UESHIBAand P. DEL BALZO Division of Pediatric Endocrinology, The New York Hospital-Cornell Medical Center, New York, NY 10021, U.S.A.
(Received 3 July 1991)
Summary--Recent reports have thrown doubt on the role of measurements of plasma 5~t-androstane-3~t,17fl-diol glucuronide (3~t-diolG) as a marker of peripheral androgen metabolism in women with polycystic ovarian syndrome and idiopathic hirsutism. It has been suggested that a plasma profile of C~9 steroid glucuronides may be more informative. While preliminary data indicates that both 3~t-diolG and androsterone G (ADTG) may arise from adrenal steroid precursors, there have been no reports of C~9 steroid glucuronides in women with non-classical, or late-onset congenital adrenal hyperplasia (NC-CAH), who constitute a significant proportion of the hirsute female population. We therefore measured plasma levels of 3~t-diolG, ADTG and dihydrotestosterone G (DHTG) before and following a standard Cortrosyn test in 15 symptomatic and 3 asymptomatic NC-CAH patients, 5 heterozygote carriers for 21-hydroxylase deficiency (NCHETS) and 18 normal women. The effects of chronic glucocorticoid (GCR) therapy (> 3 months) on the C~9 steroid glueuronide profile in the symptomatic patients was also investigated. Baseline plasma levels of all 3 glucuronides were significantly (P < 0.001) higher in symptomatic patients compared with either normals or NCHETS. However, the order of discrimination was A D T G > 3~t-diolG > DHTG. There were no significant differences between steriod glucuronide levels for NCHET and normal women and the C~9 steroid glucuronide concentrations for the asymptomatic NC-CAH patients were > 2 SD above the normal means. Moderate clinical improvement was observed in all patients receiving oral GCR therapy and was accompanied by approx. 80% suppression of the plasma levels of all 3 Ct9 steroid glucuronides. This contrasts with a mean suppression of androstenedione of only 50%. However, plasma levels of the C~9 steroid glucuronides were not significantly increased in response to a short ACTH stimulation test. This may be explained by the fact that the androgen glucuronides are thought to be peripherally formed metabolites derived from unconjugated glandular secreted androgen precursors and thus their synthesis at 60 rain following adrenal stimulation may lag substantially behind that of their respective precursors. There were significant linear correlations between the levels of all 3 glucuronides, but neither correlated with Ferriman--Gallway scores, body mass index or 17-hydroxyprogesterone levels. We therefore conclude that (1) plasma levels of 3~-diolG, ADTG and DHTG are neither related to hirsutism as previously suggested, nor body weight (in normal weight individuals) and since they are elevated in asymptomatic NC-CAH patients they may therefore only reflect adrenal contribution to circulating androgens and not peripheral androgen action per se; (2) a plasma profile of Cj9 steroid glucuronides may serve as an excellent index of the effectivenesss of treatment in hyperandrogenic disorders of primarily adrenal etiology; (3) the heterozygosity of NC-CAH does not appear to be expressed in terms of a gluco-conjugated steroid profile intermediate between that of homozygous unaffected and affected individuals; and (4) there may be a common pathway of C~9 steriod glucuronide metabolism, the precursors of which are likely to be of exclusively adrenal origin in non-obese women.
INTRODUCTION
It is well k n o w n that a relatively mild defect o f adrenal steroidogenic enzyme activity, i.e. 21*To whom correspondence should be addressed at: Department of Medicine, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, England.
hydroxylase, l l]~-hydroxylase, or 3~-hydroxysteroid dehydrogenase which manifests during either peri- or post-pubertal life, m a y constitute a significant cause o f hyperandrogenism in w o m e n [ I - 5 ] . Non-classical (NC) congenital adrenal hyperplasia ( N C - C A H ) due to mild 21-hydroxylase deficiency (21-OH) results in
211
212
C.B. WttORWOODet al.
markedly increased plasma androgen levels as a consequence of the accumulation of the C21 adrenal steroids above the 21-hydroxylase step in the metabolic pathways leading to cortisol, and is the most common autosomally recessive inherited disorder known to man [3, 5]. This allelic variant of classical CAH [1, 2] has the specific HLA associations of NC-21OH, B14 DR1, which are distinct from those frequently linked to the classical condition [6]. The clinical presentation of NC-21OH is extremely heterogenous. Symptoms may manifest at any age and include premature pubarche, accelerated growth velocity, advanced bone age and ultimate short stature in pre-adolescent individuals; and hirsutism, oligomenorrhea, acne, alopecia and infertility in adult women [2, 7, 8]. Furthermore, there is a small proportion of women whose endocrinological profile is consistent with NC-21OH and yet they are asymptomatic, often having been diagnosed as a consequence of both genetic and hormonal family studies of classical patients [9]. Clearly, those patients who manifest symptoms later in life represent a major proportion if not the entire population of asymptomatic juvenile NC-CAH subjects. The considerable qualitative and quantitative symptom variability between individuals presenting with hyperandrogenic disorders, such as NC-CAH and polycystic ovarian syndrome (PCOS), in the presence of similar circulating androgen levels, has led to the search for a suitable plasma marker of peripheral androgen metabolism and target organ sensitivity. The potent active androgen, dihydrotestosterone (DHT) and 5~-androstane-3~,17fl-diol (3~-diol) are both 5~-reduced metabolites of circulating testosterone (T), but plasma measurements fail to adequately distinguish the underlying metabolic abnormalities associated with the varied clinical presentation in hirsute women from the level of peripheral androgen action for non-hirsute women [10-12]. Plasma levels of free T may offer improved discrimination but this remains unreliable [13]. Several recent reports have suggested that plasma levels of 3~-diol glucuronide (3ct-diolG) and, more recently, androsterone glucuronide (ADTG) may reflect peripheral androgen metabolism in hirsute women with either PCOS or idiopathic hirsutism (IH)[10, 14-19]. However, Scanlon et al. [20] and others[21,22] were unable to find any significant differences between plasma 3~t-diolG levels in hirsute vs asymptomatic
women with PCOS or between normal and IH women [18, 22] thereby throwing some doubt on the significance of plasma 3~-diolG measurements. Plasma concentrations of ADTG, on the other hand, were found to be several fold those of 3~-diolG and significantly elevated both in symptomatic women with PCOS [19] and in IH patients [18] compared with non-hirsute women with PCOS. It has therefore been suggested that a C19 steroid glucuronide profile may be more informative with respect to an analysis of peripheral androgen action and furthermore may indicate the preferred metabolic pathways [18], which may be relevant in assessing appropriate specific treatment. Recent evidence suggests that both 3~tdiolG and ADTG may be derived almost exclusively from adrenal androgen precursors [18, 20, 23, 24]. We therefore decided to measure the levels of these steroid glucuronides and those of DHTG, a likely precursor of circulating 3~t-diolG, in blood obtained from (1) untreated symptomatic women with NC-21OH; (2) asymptomatic women with NC-21OH; (3) women diagnosed as being heterozygote carriers for NC-21OH; and (4) normal women. In addition, the effects of a short ACTH [1-24] test and long-term glucocorticoid therapy upon the plasma steroid glucuronide profile were also studied.
MATERIALS AND METHODS
Subjects
The diagnosis of NC-21OH is well established. Eighteen hyperandrogenic women were diagnosed as suffering from this disorder either through initial presentation at our clinic with symptoms of androgen excess (n = 15) or through genetic and hormonal family studies (n = 3). A Cortrosyn test was performed and a 24 h urine collection obtained. The diagnosis was made on the basis of a significantly elevated urinary pregnanetriol excretion and baseline and post-Cortrosyn 17-hydroxyprogesterone (17OHP) and androstenedione (A4) levels within the NC-21OH domain of a nomogram defined by our previous studies. Subsequent HLA typing and DNA analysis confirmed the diagnosis [25]. A Cortrosyn test was also performed on 18 normal women and 5 compound heterozygote carriers for NC-21OH (NCHETS), who were either the mother or sibs of affected individuals.
S t e r o i d g l u c u r o n i d e s in N C - C A H
All the women in the study were 14--45 years of age. There was no significant age difference between any of the groups. The FerrimanGallway (FG) score [26] for the normal women, asymptomatic NC-21OH and NCHETS was < 5, while that for the NC-21OH patients was 17 ___4. The body mass index [BMI: weight (kg)/height (mE)] was normal (
