This review aims to provide an update of recent areas of discussion in the literature as well as a reminder of clinical management options.
Polycystic ovarian syndrome (PCOS) remains one of the most common disorders we all deal with in one of its many forms on a weekly basis. For our understanding of such a common condition to still be evolving in all areas (diagnosis, aetiology, investigation and management) is evidence of its complexity. Prevalence depends on diagnostic criteria used and the population studied, but is about 6–12 per cent of reproductive-aged women.
The Rotterdam consensus criteria remain the most widely used diagnostic tool for PCOS. Patients must exhibit two of the following: biochemical/clinical signs of hyperandrogenism, oligomenorrhoea (fewer than nine periods/year) with no other cause and PCO morphology on ultrasound.1 Ethnic variation is common. Asian women have fewer hyperandrogenic symptoms, but significant metabolic complications while Middle Eastern and Mediterranean women have more hirsutism.2
There is current uncertainty regarding the diagnosis of PCOS in adolescents. A Danish study described PCO morphology as per the Rotterdam criteria in 68 per cent of women age 19–21 years in community-based screening, suggesting we may have been over-interpreting ultrasound findings in adolescents.3 Acne can be a normal phenomenon of adolescence. Care is needed in timing from menarche when diagnosing oligomenorrhoea. Most adolescents have a regular cycle two years postmenarche and 90 per cent of adolescents who remain oligomenorrhoeic at four years will have pathology.4 Despite these concerns, there are no clear criteria for diagnosis in adolescents as yet. The Endocrine Society of America suggests hyperandrogenism and anovulatory symptoms abnormal for gynaecological age.5 Other experts argue that ovarian enlargement should also be required.6 Pragmatically, a careful explanation of these concerns to the patient with a plan to re-evaluate the diagnosis as adolescents move into their early 20s is sensible. Clinical aims of a normal body mass index (BMI) and control of both menstrual and hyperandrogenic symptoms can still be achieved. It is possible in the future that anti-Müllerian hormone maybe a diagnostic tool in this setting.7
In lean women with oligomenorrhoea, the differential diagnosis can be between hypothalamic amenorrhoea (HA) and PCOS. The two conditions are the opposite of each other (see Table 1). A careful history is vital – periods will improve with energy deficit in PCOS and deteriorate with HA. There is potential to have a mixed picture of pathology.8 In women with a HA diagnosis who change their lifestyle and have menstrual resumption, periods may remain irregular and androgenic symptoms become more prominent. It is possible that these are the women with PCO morphology on original ultrasound, but longitudinal studies are needed.
While a detailed description of the pathogenesis of PCOS is beyond the scope of this article, insulin resistance is likely to be central to aetiology.9 A lifestyle assessment is mandatory for every woman presenting with PCOS. There is no good data suggesting any particular dietary regimen is superior to another. Sustainable long-term energy deficit is key to achieving weight loss and a multidisciplinary approach with dietician and psychology support is helpful. Fitness and achievement of a normal range BMI is without side effects and will help all areas of management. Remember ethnic appropriate BMI ranges. A BMI of 20–25 is appropriate for European people, but people from the Asian and Indian subcontinents should have a healthy BMI range of probably about 18–23.10
The combined contraceptive pill
For women who do not require immediate fertility, the combined oral contraceptive pill (COCP) remains the cornerstone of pharmacological management of PCOS. It provides endometrial protection, contraception and androgenic symptom control. The risk of venous thromboembolism (VTE) needs to be carefully considered. This risk is cumulative and other risk factors may encourage other medication choices – obesity, family history of VTE or smoking. Data is observational and mixed. Broadly absolute risk is 3/10 000 women years (WY) for non COCP users, 6/10 000 WY for COCP users. Levongesterol-containing COCPs carry approximately half the risk that drosperidone [rate ratio 1.64 (1.27–2.10)] and cyproterone acetate [rate ratio 1.88 (1.47–2.42)] containing COCPs do.11 While the COCP may cause a small change in both insulin sensitivity and lipid profile, the clinical effect of this is marginal and outweighed by the multiple benefits the COCP provides for PCOS women.12 Long-term outcome studies of COCP use have been extremely reassuring in terms of malignancy and cardiovascular risk. Hannaford and colleagues present a 40-year prospective cohort study following over 45 000 women.13 Ever COCP users had a significantly lower mortality from all causes (relative risk 0.88 [CI 0.82-0.93]) than never COCP users. Other studies support these findings and suggest we can endorse COCP use with reassurance in most women.
|Characteristic||Polycystic ovarian syndrome||Hypothalamic amenorrhoea|
|Menstruation||>Improves with leanness and fitness||>Improves with weight gain and stress relief|
|Clinical symptoms||Hyperandrogenic symptoms||If severe lanugo hair|
|Hormone tests||Elevated T, PRL, LH normal E||Low LH, E, T and sometimes low FSH|
|Ultrasound findings||Normal/thick endometrium, PCO morphology||Thin endometrium|
|Bone density||Normal||Lowered often in the spine relative to the hip|
The androgenic symptoms of acne, hirsutism and alopecia can be particularly distressing. Drosperidone or cyproterone acetate containing COCPs are very effective for androgenic symptom control. In women unable to take a COCP or if more potent androgenic control is required, options include:
- Spirinolactone 100–200mg/day. Renal function and potassium should be monitored if using with a drosperidone COCP.
- Cyproterone acetate 50–100mg ten days/month. Liver function tests should be monitored.
- Flutamide 250mg oral dosing. Liver function tests must be monitored – liver failure has been reported.
- Finasteride 2.5mg oral dosing may be helpful for androgenic alopecia.
- Vaniqa – eflornithine hydrocholride a cream applied twice daily to prevent new hair growth.
It is important to explain acne control will take three-to-four months of medication, hirsutism control 9–12 months and alopecia 12–18 months. Contraception must be used in combination with any of these medications and in the absence of the COCP this could include a Mirena, Cerazette, Depo Provera or Jadelle.
Metabolic status should be assessed at diagnosis and every one-to-two years thereafter depending on other risk factors. Assessment should include BMI, waist circumference, blood pressure, HbA1c, liver function tests, fasting lipids and glucose. An oral glucose tolerance test (OGTT) should be done at diagnosis in all women and repeated annually in those with other risk factors – BMI>25, age >30, family member with type 2 diabetes or a history of gestational diabetes. An OGTT is the best screening test for impaired glucose tolerance. There is interest in whether women with all three Rotterdam criteria are at higher risk for metabolic syndrome.14 Gastric bypass surgery has been found to be effective in managing PCOS symptoms for those that qualify.15 There is good data supporting pregnancy safety post-bypass surgery after a waiting period of 12–18 months.16
PCOS is a hyperoestrogenic condition with comparatively low levels of progesterone secondary to annovulation. Therefore, women are considered to be at risk of endometrial hyperplasia and carcinoma; however, we do not have good trials supporting increased pathology or best management.17 The number of periods required per year to provide reassurance is not well described. If women are consistently having periods less frequently, then six-weekly endometrial protection should be advised. Again the COCP is most commonly used and has good evidence for endometrial protection.18 Other options include the Mirena IUD, Provera 10mg used for 21 days per month, Depo Provera or Jadelle. Despite reassuring data for these modalities, dysfunctional bleeding with any medication should still be assessed if other risk factors such as age or obesity are present.19 Metformin is second-line treatment for endometrial protection as there are no good data supporting its use for this indication.
Women with PCOS have a higher risk of gestational diabetes and pre-eclampsia during pregnancy20; risks which are minimised with achieving a non-smoking status and normal BMI pre-conception. For women with a normal BMI, clomiphene citrate remains first-line management for annovulatory PCOS. It is extremely effective 60 per cent cumulative pregnancy rate over six months21, cost effective and relatively non-invasive compared to other fertility treatments. Metformin is now only infrequently used to assist fertility. If clomiphene is unsuccessful in achieving pregnancy, further options include ovarian diathermy, ovulation induction with FSH or in vitro fertilisation.
Adult women with PCOS have higher rates of depression standardised mean difference (SMD) 0.6 (95 per cent CI 0.47- 0.73), anxiety SMD 0.49 (95 per cent CI 0.36-0.63) and eating disorders than control groups.22 Adolescents with PCOS have not been as extensively studied and the data are not so clear.23 However, all women with PCOS should be screened for mood change. A high level of suspicion for eating disorders should be held, especially as women lose weight. Early recognition and psychology support is likely to be beneficial in all of these areas.
Place of metformin
Metformin has been the gypsy of PCOS management for the last 20 years. The mechanism of action of metformin suggests that it should be widely used in management; unfortunately, both trial data and clinical experience have proved disappointing.
There are good data supporting metformin use in impaired glucose tolerance or type 2 diabetes.24 For other manifestations of PCOS data is either not available or discouraging. About 50 per cent of women will have a more regular menstrual cycle on metformin than placebo.25 While this is a second-line option for endometrial protection it maybe helpful for those of low risk, such as adolescent girls who are not sexually active and not wanting the COCP. Metformin is likely to help with conception, but it is clearly not as effective for live birth rate as clomiphene.26 Metformin is a weak anti-androgen, but is safe to conceive on so is an option to treat acne in women actively trying to conceive.
Weight loss with metformin is widely discussed on the internet. Some studies do suggest a modest benefit. Gokcel and colleagues showed a 9–14 per cent reduction in BMI after six months of either metformin (850mg twice a day), sibutramine or olistat; with no significant difference between the three pharmacological options.27 If using metformin for this indication, lifestyle counselling must also be given. Higher doses of metformin maybe more effective, but doses are often dictated by patient tolerance. There is some interest in using metformin prophylactically in young girls with both premature adrenarche and potential insulin resistance, but as yet there are no studies to support this approach.28
Ideally, a multidisciplinary approach is indicated for all women with PCOS. The general practitioner, endocrinologist, gynaecologist, psychologist and dietician all have important roles to play. Resource scarcity means role blurring occurs and therefore an appreciation of broader issues with PCOS management is important for us all.