Hirsutism and virilization

Published on 09/03/2015 by admin

Filed under Obstetrics & Gynecology

Last modified 22/04/2025

Print this page

This article have been viewed 1710 times

CHAPTER 26 Hirsutism and virilization

Julian Barth

Introduction

Hair follicles are stimulated by androgens to either enlarge on the face and body or to shrink on the vertex of the scalp. These changes occur in the female after puberty and probably develop to a small extent in all women. A small number of women develop excessive changes in hair, prompting advice from beauty therapists or physicians. The heaviness of facial or body hair growth that is necessary for a woman to consider herself to be hirsute is dependent on racial, cultural, social and economic factors. Although the pattern of hair growth is similar in all races, there is a variation in the heaviness of growth between ethnic groups. Furthermore, Shah (1957) in India and Lunde and Grottum (1984) in Norway have both stressed that women are more likely to consider themselves to be hirsute if they have significant growth on the upper lip, chin, chest and upper back.

The problem with defining hirsutism is that all measurements are subjective (Barth 1997). Large studies of unselected women in the USA (DeUgarte et al 2006) and Scandinavia (Lunde and Grottum 1984) have shown that it is impossible to separate hirsute from non-hirsute women on the basis of hair scoring scales, as women with similar hair scores may or may not consider that they have excessive hair growth. The problem with physician scales is that they do not consider the effect of hair localization. Is a women with facial hair but no body and limb hair more or less hirsute than a woman with a hairless face but a hairy trunk and limbs?

The standard scoring scale designed by Ferriman and Gallwey (1961) (Figure 26.1) defined 1.2% of females as hirsute, but this was based purely on mathematical grounds (i.e. 2 standard deviations from the mean). This contrasts with women’s perspectives. A study of 400 unselected students, 60% of whom were Welsh, showed that 9% considered themselves to be disfigured by their facial hair growth (McKnight 1964). A study of unselected women in the USA found that approximately 20% of women had modified Ferriman and Gallwey scores greater than 3, and of these women, 70% considered themselves to have excessive facial and body hair. Furthermore, women with hirsutism give their hair growth a higher score than their physicians (Kozloviene et al 2005).

Figure 26.1 The hirsuties scoring scale designed by Ferriman and Gallwey. The body is dived into 11 zones as illustrated, and each zone is scored 0–4 giving a maximum score of 44. Ferriman and Gallwey excluded the lower arm and leg, and defined hirsuties as a score of >8. Unfortunately, there is considerable interobserver variation in scoring, so all investigators need to standardize their individual score rates.

The subjectivity in the concept of hirsuties can be simplified by accepting that any woman who complains of excessive hair is hirsute. It then follows that the physician’s role is to determine whether invasive investigations (see below) are required, and what sort of therapy is most appropriate for the degree of hair growth.

Hirsuties or hypertrichosis?

Hirsutism is a purely clinical diagnosis which is based on the growth and pattern of hair on the face and body of women. The pattern of hair growth in hirsute women is assumed to follow a similar pattern to that which develops on males after puberty. Hair first appears on the upper lip, followed by the chin and cheeks, and then the lower legs, thighs, forearms, abdomen, buttocks, chest, back, upper arms and shoulders. The hair shafts themselves are coarse, curly and tend to rise above the skin surface, rather than lying flush with the skin as in hypertrichosis.

Hypertrichosis needs to be differentiated as it is not androgen mediated and therefore does not respond to antiandrogen therapy. The onset of hair growth is not related to puberty; there is no hair growth pattern but there is a uniform diffuse growth of hair over the body. The individual hair shafts tend to be finer and lie flat against the skin, rather than being coarse, curly and rising above the skin surface. Hypertrichosis may be congenital or acquired due to drug therapy and some metabolic disorders (Table 26.1).

Table 26.1 Aetiology of hirsuties and hypertrichosis

Causes of hirsutism	Causes of hypertrichosis
Ovarian, e.g. polycystic ovary syndrome, ovarian tumours Adrenal, e.g. congenital adrenal hyperplasia, adrenal tumours Cushing’s syndrome Gonadal dysgenesis Androgen therapy Valproate therapy for epilepsy Obesity (due to consequent PCO) Prolactinoma (due to consequent PCO) Stress Idiopathic

Drug therapy, e.g. cyclosporin, diazoxide, minoxidil, penicillamine etc.

Congenital disorders, e.g. hypertrichosis lanuginosa and other syndromes often associated with mental retardation

Acquired hypertrichosis lanuginosa as a paramalignant condition

Metabolic disorders, e.g. lipodystrophy, mucopolysaccharidoses, malnutrition

PCO, polycystic ovaries.

Investigation

A diagnosis of hirsutism is based purely on the patient’s history and clinical examination. The purpose of investigation is to determine if there is any significant underlying condition which will influence further management. The majority of patients will have underlying polycystic ovary syndrome, and any investigative strategy should be designed to select those subjects who have a significant endocrinopathy that requires specific therapy.

Diagnostic approach to the hirsute woman

The history of the duration and severity of hirsutism is helpful if there is suspicion of a serious underlying cause:

• longstanding mild hirsutism with regular cycles and no features of systemic virilization; these women require no further investigation;

• longstanding moderate hirsuties with or without menstrual cycle disturbances. Most of these patients will have polycystic ovaries. Further investigation is only required if serum testosterone is elevated or if the patient seeks therapy for another symptom related to hyperandrogenism, such as infertility; and

• recent-onset hirsuties and virilization, or very severe hirsuties with a long history. Features of systemic virilism include: cliteromegaly, deepening of the voice, recent-onset temporal balding, increasing muscle bulk and decrease in breast size. These women require a comprehensive endocrine evaluation for an androgen-secreting tumour (Figure 26.2).

Figure 26.2 Diagnostic approach for a woman complaining of excess facial and/or body hair. PCO, polycystic ovaries; LH, luteinizing hormone.

What is the evidence for this approach?

There have been two large studies to support this approach. Both confirmed the very low prevalence of serious endocrine disease, and it is clear that with the evidence of clinical case histories, these conditions are largely apparent on history and clinical examination.

Moran et al (1994) reported on 250 premenopausal (<38 years) women presenting with various symptoms who were noted to be hirsute. Ninety-six percent were diagnosed with polycystic ovaries, idiopathic hirsuties or obesity; 2% were diagnosed with congenital adrenal hyperplasia (CAH); and 1.6% were either iatrogenic (one case) or due to an ovarian tumour (two cases) or Cushing’s syndrome (one case).

O’Driscoll et al (1994) reported a series of 350 consecutive patients who had presented with either hirsuties or androgenic alopecia. They identified eight women from this series with clear endocrine diseases; two had pituitary tumours (prolactinoma and acromegaly). The others were identified by a grossly elevated serum testosterone (>5 nmol/l). In total, 13 women had elevated testosterone; six had polycystic ovaries, two had late-onset CAH, one had virilizing CAH that had been ignored in childhood, one had an adrenal carcinoma, one had a postmenopausal Leydig cell tumour, one had corticosteroid 11-reductase deficiency, and the diagnosis in one hyperandrogenized postmenopausal woman was not elucidated. On the basis of the clinical details given in the report, only two diagnoses of late-onset CAH would have been missed by clinical evaluation alone (one had regular cycles and two children, the other had a long history of oligomenorrhoea; both had polycystic ovaries on ultrasound examination).

Arguably, the most important purpose for investigation is to identify those women with androgen-secreting tumours, since they require different therapy. Derksen et al (1994) reported a series of two adrenal adenomas and 12 carcinomas in which hirsutism was the presenting symptom. The women with adenomas were described as being severely virilized. Six of the women with carcinoma had clinical signs of Cushing’s syndrome; of the remaining six cases, four were severely virilized and the other two women presented with abdominal pain.

Functioning ovarian tumours which secrete androgens and therefore cause virilization are rare and represent only 1% of all ovarian tumours. In these cases, hirsuties is a nearly universal feature. Amenorrhoea develops rapidly in all premenopausal patients, and systemic virilization with alopecia, cliteromegaly, deepening of the voice and a male habitus develops in approximately half of patients. Meldrum and Abraham (1979) reviewed the literature of 43 women with virilizing ovarian tumours; seven had a serum testosterone level below 7.0 nmol/l but all were clinically virilized, and one 65-year-old woman was reported as not virilized but her serum testosterone level was more than 12 nmol/l.

In conclusion, the vast majority of hirsute women have polycystic ovaries, and those women with androgen-secreting tumours can be identified by clinical evaluation and a single serum testosterone estimation (Figure 26.3).

Figure 26.3 Diagnostic approach for a woman with features of systemic virilism. U/S, ultrasound; CT, computed tomography; MRI, magnetic resonance imaging; NAD, no abnormalities detected.

How Useful Are Endocrine Tests?

The laboratory findings in reports of hirsute women demonstrate considerable heterogeneity. This may be due to differences in diagnostic criteria for the studies; for example, hyperandrogenaemia, acne and/or hirsutism, or women with polycystic ovaries/menstrual disturbances. A further factor may be the degree of hair growth necessary to define whether a woman is considered to be hirsute (Barth et al 2007). A third point has been proposed by Toscano et al (1983) who suggested that the endocrine abnormalities are dependent on the length of the history of hirsutism. They posed the question ‘is hirsutism an evolving syndrome?’, and whilst their interpretation of the findings may be argued, it is clear that (i) a degree of hirsutism and/or oligomenorrhoea is necessary before hormonal abnormalities may be present, and (ii) not all hirsute women have abnormal circulating hormones.

Investigations to diagnose polycystic ovaries

A diagnosis of polycystic ovaries was originally made on the basis of clinical signs and gross anatomical features of the ovaries; indeed, the majority of cases can be diagnosed on the basis of hirsuties and oligo-/amenorrhoea. Confirmation of the clinical diagnosis is made variably by ultrasound imaging or endocrine tests. The choice of test is largely philosophical; imaging makes a diagnosis that matches the anatomical description of the condition, whereas endocrine tests can, at best, only act as surrogates.

Ovarian imaging compares extremely favourably with anatomical appearance in those subjects who have had abdominal laparoscopy, whereas the use of gonadotrophins is less well established as a diagnostic tool. Approximately 50% of women with polycystic ovaries have an elevated concentration of luteinizing hormone (LH), and 40% of women with elevated LH have polycystic ovaries. The use of gonadotrophins is further complicated by assay methodology. In view of these factors, the latest guidelines to diagnose polycystic ovaries have abandoned their use (Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group 2004). The over-riding problem is that approximately 10% of women with ultrasound-diagnosed polycystic ovaries will have normal endocrine tests, however many are performed. This may be due to the relationship between body weight and the dynamics of testosterone production (Figure 26.4).

Figure 26.4 Although serum testosterone is usually reported as slightly elevated in women with polycystic ovaries (PCO) or hirsuties, it can be seen from this redrawing of the original data of Bardin and Lipsett (1967) that testosterone production rates are closely related to body mass irrespective of the patient’s phenotype. IH, idiopathic hirsuties.

Data redrawn from Thomas PK, Ferriman DG. Variation in facial and pubic hair growth in white women. Am J Phys Anthropol. 1957 Jun;15(2):171–180, Wiley.

Most recommendations suggest a battery of androgen tests, but this is probably still based on the unreliability of testosterone assays (Rosner et al 2007). This should become less of a problem as measurement by tandem mass spectrometry becomes more routine. A single testosterone measurement taken in the morning during the early follicular phase measured with a reliable assay is probably all that is required in the first instance (Martin et al 2008).

Short Synacthen test for late-onset congenital adrenal hyperplasia

The incidence of late-onset CAH in women with hirsutism is variously reported at 1–12% depending on the source of patient referral. Short Synacthen testing has been used extensively as a research tool, but its use in routine practice cannot be recommended. Firstly, there is no sharp cut-off value of 17-hydroxyprogesterone following Synacthen that can be used to define late-onset CAH. Secondly, treatment for hirsute women with late-onset CAH should not be influenced by the diagnosis since antiandrogen therapy with cyproterone acetate (CPA) is more effective than corticosteroids (Spritzer et al 1990). Moreover, corticosteroids aggravate insulin resistance and, theoretically, worsen the hyperandrogenism. Finally, there does not appear to be a need to identify women with late-onset CAH since they do not seem to need protection against stress-related corticosteroid deficiency.

Evaluation of metabolic factors

Polycystic ovary syndrome is now recognized to be a systemic disorder underpinned by insulin resistance. Insulin plays a key role as an amplifier of gonadotrophin-stimulated ovarian androgen production, and also as a growth factor for hair follicle. The effects of hyperinsulinaemia include android obesity, glucose intolerance, hyperlipidaemia and possibly impaired fibrinolysis. The lipid profiles show the pattern associated with patients who have ischaemic heart disease, namely raised triglycerides and low high-density-lipoprotein cholesterol, and hirsute women undergoing coronary angiography have more pronounced arterial disease than non-hirsute women (Wild et al 1990). Moreover, long-term studies of women with polycystic ovaries have shown an increase in diabetes and hypertension, and the central obesity phenotype in women carries an increased risk of cardiovascular disease (see Chapter 18, Polycystic ovary syndrome, for more information).

These metabolic factors are already well established during adolescence, as evidenced by the changes in bony structure. These must have been completed by the time of epiphyseal fusion at the end of the second decade. This was clearly demonstrated by Ferriman et al (1957) who noted that hirsute women had increased shoulder width and reduced bi-iliac width compared with non-hirsute women, and that these differences were greater in those women who noted that their hirsutism began during adolescence.

Recommendations for the investigation of hirsute women

Most women with hirsuties are either seeking reassurance that they are not changing sex or are seeking advice regarding therapy for their hair. Measurement of a single serum testosterone during the early follicular phase may be a useful therapy for the former, and is a useful first-line indicator of a serious endocrine disease, although, as discussed above, women with the latter will be identified by history and clinical examination. There seems to be no benefit to be gained by extensive endocrine tests in uncomplicated cases of hirsutism, since the decision to treat is not based on endocrine test results.

Since there are serious adverse metabolic factors in both lean and obese women with polycystic ovaries, it would be prudent to screen for impaired glucose tolerance by a fasting glucose and to measure a fasting plasma lipid profile.

Postmenopausal women

A different strategy is required for postmenopausal women because, unlike younger women, they cannot be stratified on the basis of menstrual pattern. However, as women age, the pattern of facial and body hair changes (Figure 26.5); hair on the face increases whilst it involutes elsewhere. Hair growth on the chest, shoulders and upper back can be considered indications for further investigation. Hormonal changes in hirsute postmenopausal women are minor with small increases in testosterone compared with appropriate controls, so an elevated testosterone more than twice the upper limit of age-related normal or approximately 3.0 nmol/l should also be considered as a trigger for further investigation. Postmenopausal women with unexplained hirsuties have normal lipids, but women with ovarian hyperthecosis should be investigated further since they appear to have a very high incidence of glucose intolerance and vascular disease (Barth et al 1997).

Figure 26.5 Differences in facial and abdominal hair in women aged 15–85 years. Data were collected on 584 women admitted to hospital. Hair was graded 0–4 at each site, and scores of 3 or 4 were considered to represent significant hair growth for this graph.

Redrawn from Bardin CW, Lipsett MB, Testosterone and Androstenedione Blood Production Rates in Normal Women and Women with Idiopathic Hirsutism or Polycystic Ovaries, Journal of Clinical Investigation, 1967; 46: 891–902.