Search

Polycystic ovary syndrome (PCOS) may be loosely defined as unexplained hyperandrogenism, with variable degrees of cutaneous symptoms, anovulatory symptoms, and obesity. The vast majority of patients with the full-blown Stein-Leventhal syndrome have functional ovarian hyperandrogenism (FOH). However, FOH often occurs without the LH excess or polycystic ovaries of classic PCOS. Functional adrenal hyperandrogenism (FAH) is often found in the syndrome, but it is less closely associated with anovulatory symptoms than is FOH. The vast majority of FOH seems to arise from abnormal regulation (dysregulation) of ovarian androgen secretion. This typically is due to escape from desensitization to luteinizing hormone (LH); this appears to occur because of a breakdown in the processes that normally coordinate ovarian androgen and oestrogen secretion so as to prevent hyperoestrogenism. Similar dysregulation of adrenal androgen secretion in response to ACTH seems to account for most FAH. Dysregulation of androgen secretion may affect the ovary alone (isolated FOH), the adrenal alone (isolated FAH), or both together. Modest insulin resistance is common in PCOS/FOH, and the resultant hyperinsulinaemia is a major candidate as the cause of the dysregulation. The hyperinsulinaemia may arise from either 'nature' (genetic defects) or 'nurture' (exogenous obesity). Although hyperinsulinaemia alone does not have an obvious effect on steroidogenesis, it may act in genetically predisposed women as a 'second hit' to unmask latent abnormalities in steroidogenesis. The ovary, the adrenal cortex, and several other organs paradoxically function as if responding to the hyperinsulinaemic state in spite of resistance to the effects of insulin on glucose metabolism. PCOS should be viewed as an early manifestation of a hyperinsulinaemic condition that will predispose to cardiovascular and metabolic complications later in life. A subset of PCOS patients appear to have not only insulin resistance but also beta-cell secretory dysfunction, which may indicate a relationship of the disorder to NIDDM. The fundamental genetic defects remain to be elucidated.

Hyperandrogenism must be considered in any girl with premature pubarche, unusual acne, hirsutism, or androgenetic alopecia. An association with menstrual irregularity or obesity should raise the index of suspicion. The most common causes of hyperandrogenism presenting in a teenage girl are functional ovarian hyperandrogenism, one manifestation of which is polycystic ovary syndrome, and functional adrenal hyperandrogenism, which usually seems to be due to an exaggeration of adrenarche. The plasma-free testosterone is a more sensitive indicator of hyperandrogenism than is the total testosterone concentration. The pattern of response of plasma free testosterone, DHEAS, and cortisol to dex-suppression testing can be diagnostic of the source of androgen excess. Treatment, including oral contraceptives, low-dose glucocorticoids, and antiandrogens, should be chosen according to the patient's symptoms and source of androgens and should be combined with traditional therapy for the specific dermatologic disorder.

Nonclassical 3 beta-hydroxy-delta 5-steroid dehydrogenase (3 beta-HSD) deficiency type of congenital adrenal hyperplasia has been hypothesized to occur in as many as 10-40% of hirsute women, based on the adrenal steroidogenic responses to ACTH. However, diagnostic criteria for this "late-onset" 3 beta-HSD deficiency are not clearly established. Among 40 successive hyperandrogenic women undergoing evaluation of adrenal steroidogenic responses to ACTH, 8 had responses suggestive of 3 beta-HSD deficiency. Since 3 beta-HSD is present in both the ovary and adrenal, we attempted to document the defect in the ovary by stimulating their ovarian function with a gonadotropin-releasing hormone agonist test using nafarelin (6-D-[2-naphthyl]alanine-gonadotropin-releasing hormone). The eight hirsute women had steroid responses to ACTH suggestive of 3 beta-HSD deficiency, namely, the values of the delta 5-steroids, 17-hydroxypregnenolone and dehydroepiandrosterone, 30 and 60 min after ACTH in each hirsute woman were greater than 2 SD above the normal mean. Seven of the eight hirsute women had at least one elevated delta 5/delta 4-steroid ratio; however, only three of the hirsute women had two abnormal ratios. Furthermore, the response of the delta 4-steroid androstenedione and the ratio of androstenedione to cortisol after ACTH were significantly increased in the hirsute women, findings not consistent with 3 beta-HSD deficiency. After nafarelin, five and six hirsute patients had elevated values of the delta 4-steroids androstenedione and 17-hydroxyprogesterone, respectively. No patient had an elevated delta 5/delta 4-steroid ratio after nafarelin. Thus, ovarian steroidogenic responses to nafarelin did not support the diagnosis of 3 beta-HSD deficiency. Rather, they are consistent in most cases with polycystic ovary syndrome due to dysregulation of 17-hydroxylase and 17,20-lyase activities. We propose that increased activity of the enzyme P450c17 alpha in the adrenal cortex is responsible for most of what is often termed late-onset 3 beta-HSD deficiency.

Obesity, oligomenorrhea, and hirsutism are frequently associated with high plasma androgen levels and/or low testosterone-binding globulin (TEBG) levels. Studied have been undertaken to determine the extent to which each of these clinical features may be related to this hormonal profile. Indexes of plasma free (unbound) androgen levels were focused upon because this fraction appears to be the biologically active portion of the plasma androgens. The hormonal profile was normal in women with either obesity or oligomenorrhea alone and abnormal in those with severe hirsutism alone. On striking new finding was that subjects with the combination of obesity and oligomenorrhea had elevated plasma total and free androgens and depressed TEBG even in the absence of hirsutism. Furthermore, the androgen levels were higher in obese oligomenorrheic women with mild hirsutism than in severely hirsute women who were not obese. Plasma estradiol concentrations were normal in these obese women. A simple explanation for elevated free plasma androgen levels in obese women who were oligomenorrheic yet had little if any hirsutism is not possible. The data are compatible with the concept that obesity is a variably expressed manifestation of slightly elevated plasma free androgen levels or that obesity is an incidental finding which somehow blunts the effect of high androgen levels on hair follicles. Regardless of the explanation, oligomenorrheic obese women must be suspected of having high androgen production even in the absence of hirsutism.

The relationship in women between mild hirsutism or acne to androgen levels has not been well-defined. We investigated this in 62 Caucasian women, aged 18 to 21 years, by relating these pilosebaceous signs to plasma free testosterone (fT), the main circulating determinant of plasma androgenicity. Women with mild hirsutism (n = 13) had a significantly elevated fT (12.7 +/- 5.5, SD, pg/mL [44.1 +/- 19.1 pmol/L]) compared to normal controls (7.4 +/- 2.7 pg/mL [25.7 +/- 9.4 pmol/L]), as did subjects with minor acne (10.7 +/- 4.25 pg/mL [37.1 +/- 14.7 pmol/L]). The most important finding was the striking variability in the relationship between pilosebaceous overactivity and fT levels. In mildly hirsute subjects plasma fT was normal in half of the subjects, and the coefficient of variation of plasma fT was about twice what one would expect from individual variability. We could not demonstrate correlations among the variables of hirsutism, acne, and plasma fT. On the other hand, among 15 women with modest elevations of plasma fT levels (up to twofold), 27% had moderate hirsutism, 40% had mild hirsutism, and 33% had none. However, four of five of the latter patients (without hirsutism) had acne. The relationship of fT to acne severity varied similarly. To define the interactions between androgens and the pilosebaceous apparatus, we propose a model in which variation in apparent skin sensitivity and the level of androgen seem to contribute about equally to the pathogenesis of mild hirsutism and acne. The clinician should suspect that hyperandrogenemia will be found in about half the women with mild cases of hirsutism, and one third with minor acne.

Hirsutism usually results from a subtle excess of androgens. As such, it is a clue to possible endocrine disturbance in addition to presenting cosmetic problems. We use the term hirsutism to mean male-pattern hirsutism--excessive growth of hair in areas where female subjects normally have considerably less than male subjects. An elevation of the plasma free (unbound) testosterone level is the single most consistent endocrinologic finding in hirsutism. The plasma free testosterone level is sometimes elevated when the total level of plasma testosterone is normal because testosterone-estradiol--binding globulin (TEBG) levels are often depressed in hirsute women. Frequent blood sampling is sometimes necessary to demonstrate subtle hyperandrogenic states since androgen levels in the blood are pulsatile and seemingly reflect episodic ovarian and adrenal secretion. The source of hyperandrogenemia can usually be determined from dexamethasone suppression testing. Those patients whose plasma free androgen levels do not suppress normally usually have functional ovarian hyperandrogenism (polycystic ovary syndrome variants). Very high plasma androgen levels or evidence of hypercortisolism, which is not normally suppressible by dexamethasone, should lead to the search for a tumor or Cushing's syndrome. Those patients in whom hyperandrogenemia is suppressed normally by dexamethasone have a form of the adrenogenital syndrome, a prolactinoma, obesity, or idiopathic hyperandrogenemia. In such patients, glucocorticoid therapy may reduce hirsutism and acne and normalize menses. The treatment of hirsutism resulting from functional ovarian hyperandrogenism is not as satisfactory; estrogen-progestin treatment is the most useful adjunct to cosmetic approaches to hirsutism in this country. However, other manifestations of polycystic ovary syndrome, such as infertility, may take precedence over hirsutism when an optimal therapeutic program is designed for many patients.

To determine the adrenal contribution to elevated plasma androgens in 31 young hyperandrogenemic women with acne and/or hirsutism, we compared their responses to ACTH with those of 14 normal women. Each subject was given a low dose (10 micrograms/m2) of synthetic ACTH-(1-24) (Cortrosyn) after administration of 1.5 mg dexamethasone the night before the test. Thirty and 60 min responses of plasma 17 alpha-hydroxypregnenolone (17-Preg), 17 alpha-hydroxyprogesterone, (17-prog), dehydroepiandrosterone (DHEA), androstenedione, 11-deoxycortisol, and cortisol were measured. Eighteen (58%) patients had increased responses of at least one 17-ketosteroid or adrenal androgen precursor. All patients had cortisol responses within the range of those of the 14 normal subjects. Nine patients (29%) had evidence of steroid biosynthetic enzyme deficiencies, either mild congenital adrenal hyperplasia or the heterozygote state; after ACTH, 4 of these patients had elevated 17-prog in the range of values in heterozygote carriers of 21-hydroxylase deficiency, 2 had elevated levels of 11-deoxycortisol compatible with 11 beta-hydroxylase deficiency, and 3 had elevated levels of 17-Preg and DHEA, suggestive of 3 beta-hydroxysteroid dehydrogenase deficiency. Another 9 subjects (29%) had 17-ketosteroid (DHEA and/or androstenedione) hyperresponsiveness to ACTH with associated elevated 17-Preg responses. As a group, their patterns suggested relatively deficient 3 beta-hydroxysteroid dehydrogenase and relatively hyperactive C lyase without impairment of cortisol secretion. This pattern resembles exaggerated adrenarche, and we postulate that these 9 patients have hyperplasia of the zona reticularis. Neither basal levels of plasma androgens (free testosterone and DHEA sulfate) nor menstrual history predicted which patients would have abnormal ACTH responses. Although 5 of 11 (45%) patients with acne alone had abnormal responses to ACTH, 10 of 14 patients with acne and hirsutism (71%) had abnormal responses to ACTH. We conclude that an adrenal contribution is found in about half of hyperandrogenemic women with acne and/or hirsutism. This adrenal androgen hyperresponsiveness is heterogeneous. Some patients may have mild forms of congenital adrenal hyperplasia. However, functional androgenic hyperresponsiveness to ACTH, which resembles an exaggeration of adrenarche, is the most common abnormality found. Such findings may provide an explanation for the clinical observation of exacerbations of acne with stress.

To investigate the basis of the hirsutism and elevated plasma dehydroepiandrosterone (DHA) and/or DHA sulfate (DHAS) in hyperprolactinemic women, we measured androgen binding parameters and an extensive profile of plasma androgens in normal (NL) and hyperprolactinemic women (HYPRL). ACTH tests and dexamethasone (dex) suppression tests were performed in subgroups. Free testosterone levels were higher in HYPRL (13.1 +/- 23.3 vs. 7.18 +/- 0.72 pg/ml; P less than 0.025), although total testosterone was comparable. This disparity was related to plasma testosterone-estradiol-binding globulin (TEBG) levels being one third lower in HYPRL (mean +/- SE, 27.4 +/- 4.0 nM) than in NL (41.2 +/- 3.7 nM; P less than 0.0125). Less striking elevations of plasma DHAS, androstenedione, and 11-deoxycortisol were found in HYPRL. Plasma total dihydrotestosterone [17 beta-hydroxy-5 alpha-androstan-3-one (tDHT)] was nearly 30% lower in HYPRL (11.2 +/- 2.6 ng/dl) than in NL (15.6 +/- 1.3 ng/dl; P less than 0.025), whereas free DHT was normal. Ratios of tDHT to precursors were lower in HYPRL (P less than 0.005). After ACTH stimulation, hyperresponsiveness of 17-hydroxyprogesterone and androstenedione were observed. Apparent adrenal enzyme efficiencies, judged from post-ACTH product to precursor ratios, were normal in HYPRL with one exception: the ratio of tDHT to total testosterone at 4 h was lower (P less than 0.05). Dex suppression normalized androgens and obliterated the abnormal tDHT to precursor ratios. These findings suggest an ACTH dependency of the abnormalities. In summary, we find that about 40% of HYPRL have an androgenic abnormality, and the most characteristic abnormality is an elevated free testosterone level (abnormal in 43%). Depressed TEBG and high DHAS levels were found with lesser frequency (19-21%). The plasma tDHT concentration was low, both in absolute terms and relative to its precursors. Dex suppressibility of the hyperandrogenemia was also observed. We postulate that PRL may exert multiple effects on steroid secretion and metabolism. Possibilities include the inhibition of the TEBG level.

We have tested the hypothesis that idiopathic hirsutism (IH) may be due to abnormality of androgen-responsive hair follicles. Because androgen metabolism within target cells is an important determinant of androgen action, we have analyzed the rates of formation and disposition of the major mediators of androgen action, testosterone (T) and dihydrotestosterone (DHT). In normal women, the pattern of androgen metabolism by growing hairs favors T predominance over DHT and inactivation of both these 17 beta-hydroxysteroids to 17-ketosteroids. This pattern results greatly from predominance of 17 beta-hydroxysteroid dehydrogenation. For example, in normal women's scalp hair, DHT disposition to 5 alpha-androstanedione proceeded at the rate of 8.6 +/- 2.0 (SEM) %/micrograms DNA/min, whereas DHT was formed from T at a rate of 0.14 +/- 0.02, and T was formed from androstenedione at a rate of 0.60 +/- 0.12, all significantly different from one another. Both the formation of 17-ketosteroids and the apparent 5 alpha-reductase activity were exaggerated in the pubic hair of men; whether these differences are site-, sex-, or androgen-related, remains to be determined. Pubic hairs tended to metabolize androgens at a greater rate than did scalp hair. This was related to the significantly greater DNA content of plucked pubic hairs, a difference unrelated to sex or androgen levels. Women with IH had heterogeneous pubic hair abnormalities. Only 1 of the 4 IH patients studied had abnormal pubic hair follicle androgen metabolism, with the greatest abnormality being an exaggerated rate of 17 beta-hydroxysteroid inactivation to 17-ketosteroids. Two of the other 3 IH cases had increased DNA content of plucked pubic hairs, a different kind of exaggeration of normal, which suggests an abnormality of hair follicle growth unrelated to androgen sensitivity. We favor the concept that IH is related to various distinct types of sexual hair abnormalities which reflect fundamental defects in the regulation of hair growth.

PSAs, with few exceptions, consist of a piliary and a sebaceous component. In androgen-sensitive areas, each has the capacity to develop into either a terminal hair follicle or a sebaceous follicle depending upon its location. Without androgen, there is no development of the sexual hair follicle or sebaceous gland. Androgens appear to promote sexual hair growth by recruiting a population of PSAs that have preset genetic sensitivity to initiate the production of terminal hairs. The site of action of androgens within the PSA is unclear. There are indications that androgens may act at more than one site in a system that requires two-way reciprocal interaction between dermal and epithelial cells for the generation of hair growth. Growth hormone appears to exert an important synergism with androgen in affecting the PSA, seemingly through the mediation of insulin-like growth factors. Hirsutism is due to an increased density of growing terminal hairs. The majority of cases of moderately severe hirsutism in women are due to hyperandrogenaemia, as are half the cases of mild hirsutism and about one-quarter of the cases of mild acne vulgaris. We advocate reserving the term idiopathic hirsutism or idiopathic acne for those patients in whom excessive growth of terminal hair or acne is not explained by androgen excess. We believe that highly variable sensitivity to androgen within the population explains both idiopathic hirsutism and cryptic hyperandrogenaemia; that is, these disorders lie at opposite ends of the normal spectrum of sensitivity to androgen. The biological basis for the variations in responsiveness of PSAs to androgens is unknown. The regression of hirsutism induced by antiandrogen treatment is characterized by the growth of hairs that are more vellus in character, i.e. smaller and less medullated.