Association between Type 2 diabetes mellitus and polycystic ovary syndrome (PCOS)

• Prospective studies provide evidence that the prevalence of both Type 2 diabetes mellitus and impaired glucose tolerance is higher in women with PCOS than in age- and weight-matched women without PCOS.
• There is evidence from a non-systematic review and small case-control and cross-sectional studies that women with PCOS are more likely to have insulin resistance, and that this is more marked if the woman is obese.
• Good evidence from prospective cohort studies indicates that women with PCOS have an increased risk of developing Type 2 diabetes or impaired glucose tolerance in middle age. Obesity increases the risk of impaired glucose tolerance and diabetes [Legro, 2001].
• Limited evidence from a prospective study indicates that women who have had gestational diabetes mellitus are more likely to have PCOS.

Screening

• These recommendations are based on expert advice from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007], a position statement from the Androgen Excess Society [Salley et al, 2007], and the opinion of CKS expert reviewers.
• It is not known whether or how often women with PCOS should be screened for impaired glucose tolerance or Type 2 diabetes.
• Three cohort studies provide evidence that fasting glucose levels alone are a poor predictor of impaired glucose tolerance and that a family history of diabetes in a first-degree relative may be significant. The evidence suggests that if a test is done, an oral glucose tolerance test is the most appropriate. However this is not always practical in primary care.
• The Royal College of Obstetricians and Gynaecologists suggest an approach to screening that does not involve performing an oral glucose tolerance test on all women with PCOS; it recommends regular (perhaps annual) fasting glucose tests in all women with PCOS but performing an oral glucose tolerance test in women who have a body mass index (BMI) greater than 30 kg/m², a strong family history of Type 2 diabetes, or a fasting glucose level of 5.6 mmol/L or greater [RCOG, 2007].
• The International Diabetes Federation suggests that if the fasting glucose level is greater than 5.6 mmol/L, an oral glucose tolerance test should be done [International Diabetes Federation, 2005].
• A position statement from the Androgen Excess Society suggests that all women with PCOS should have an oral glucose tolerance test. Most members of the Androgen Excess Society Board suggest screening all women with PCOS at least every 2 years with an oral glucose tolerance test, or more frequently if additional risk factors are present. Women with impaired glucose tolerance should be screened annually for Type 2 diabetes. A few members of the Board suggested an oral glucose tolerance test only in women with a BMI of 30 kg/m² or in lean women with additional risk factors [Salley et al, 2007].
• Taking into account the practicalities of screening in primary care, these recommendations are based on:
• The position statement from the Androgen Excess Society that all women with PCOS should have an oral glucose tolerance test [Salley et al, 2007] and the opinion from some CKS expert reviewers that all women with PCOS should have an initial oral glucose tolerance test.
• The position statement from the Androgen Excess Society that all women with impaired glucose tolerance should be screened annually for Type 2 diabetes mellitus [Salley et al, 2007].
• The pragmatic recommendation from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007], and also from some CKS expert reviewers, of annual fasting glucose measurement for all women with PCOS.
• The suggestion from the Androgen Excess Society of doing oral glucose tolerance tests more frequently than once every 2 years in women at particular risk of Type 2 diabetes [Salley et al, 2007].
• CKS has also recommended considering women who have had gestational diabetes mellitus or Asian women with a BMI greater than 25 kg/m² to be at particular risk of Type 2 diabetes; CKS expert reviewers agreed with this suggestion.

• Guidelines from the Royal College of Obstetricians and Gynaecologists make the following points about the use of these drugs in women with polycystic ovary syndrome [RCOG, 2007]:
• There is no evidence of long-term benefit.
• Evidence suggests that metformin may have short-term effects on insulin resistance in women without diabetes.
• Metformin may reduce androgen levels by about 11% compared with placebo.
• Evidence regarding metformin and reduction in body weight is conflicting.
• Women with a body mass index greater than 37 kg/m² may not respond to metformin.
• No evidence supports the use of insulin-sensitizing drugs in the prevention of cardiovascular disease.
• Evidence suggests that metformin is no better than diet and lifestyle at improving metabolic risk and progression to Type 2 diabetes.
• There has been recent concern that myocardial infarction and death are increased in women with diabetes treated with rosiglitazone.
• Metformin and the thiazolinediones are unlicensed for use in polycystic ovary syndrome.
• As there is uncertainty about the benefits and safety of insulin-sensitizing drugs, CKS recommends that they be initiated by a specialist.
• There is no expert consensus on which women should be referred for treatment with metformin.

These recommendations are based on guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007].

Risk factors for cardiovascular disease

• Limited evidence from a non-systematic review, a small cross-sectional study, and a case-control study suggests that women with polycystic ovary syndrome (PCOS) have more risk factors for cardiovascular disease than other women of the same age. In particular, they may have central obesity, hypertriglyceridaemia, and reduced high-density lipoprotein (HDL) cholesterol levels.
• There is evidence from a non-systematic review and case-control studies that women with PCOS frequently have abnormal lipid profiles, with increased triglyceride and total and low-density lipoprotein cholesterol levels. However the effect of PCOS on HDL cholesterol is controversial [RCOG, 2007].
• A systematic review found that evidence linking PCOS and hypertension was conflicting.
• Central obesity is observed in 35–60% of women with PCOS [Balen and Glass, 2005].
• Hyperandrogenism is associated with a preponderance of fat localized to truncal abdominal sites, and women with PCOS have greater truncal abdominal fat distribution (demonstrated by a higher waist-to-hip ratio).
• This central distribution of fat is independent of body mass index and is associated with higher plasma insulin and triglyceride concentrations and lower HDL cholesterol concentrations.
• Limited evidence from a small cross-sectional study indicates that women with PCOS have elevated concentrations of plasminogen activator inhibitor 1, which is a potent inhibitor of fibrinolysis, and has been linked to insulin resistance and an increased risk of thrombotic vascular events [Hopkinson et al, 1998; Kelly et al, 2000].
• Hyperhomocysteinaemia is a recognized risk factor for atherosclerosis. Limited evidence from a small case-control study suggests that plasma homocysteine levels are higher in women with PCOS than in an ethnically matched control group.

Risk of cardiovascular disease

• It is not known whether women with PCOS are at greater risk of cardiovascular disease.
• Limited evidence from epidemiological studies shows that morbidity and mortality from coronary heart disease in middle-aged women with a history of PCOS is not as high as predicted, despite their increased cardiovascular risk factors [Wild, 2002b; Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004].
• However, the cohorts of women studied so far have been relatively young (around 55 years) [Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004].
• Unknown factors in PCOS may protect the heart despite the presence of other risk factors [Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004].
• Limited evidence suggests an increased risk of stroke.
• It is not clear whether PCOS is an independent risk factor for cardiovascular disease [Wild, 2002b].

Recommendations for screening for cardiovascular risk factors

• At present, it is unclear which women should be screened for cardiovascular risk and when this should occur. These recommendations are based on expert opinion and consensus.
• Guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007] suggest that it is prudent to screen women with PCOS for cardiovascular risk factors.
• It has been suggested that only women 35 years of age or older with PCOS should be screened [Lobo and Carmina, 2000] because abnormal lipid levels in younger women usually do not require treatment [Harborne et al, 2003].

Recommendations for treatment

• Although women with PCOS are more likely to have abnormal lipid profiles, which may put them at higher risk of cardiovascular disease, it is not known whether modifying the risk factors in women with PCOS reduces cardiovascular events. No studies have been sufficiently powered to assess the effects of preventive therapy on cardiovascular morbidity and mortality in women with PCOS [Wild, 2002b].
• Although conventional cardiovascular risk calculators have not been validated in women with PCOS, guidelines from the Royal College of Obstetricians and Gynaecologists recommend that clinicians should continue to identify cardiovascular risk factors in women with PCOS and treat them according to current guidance [RCOG, 2007].
• CKS recommends that, in view of the possible increased risk of cardiovascular events, clinicians should give advice on measures to reduce cardiovascular risk.

• Guidelines from the Royal College of Obstetricians and Gynaecologists recommend asking all women with polycystic ovary syndrome whether they have symptoms of obstructive sleep apnoea [RCOG, 2007].
• Limited evidence from observational studies indicates that obstructive sleep apnoea is more prevalent in women with polycystic ovary syndrome than can be explained by obesity alone.

Advise weight loss if appropriate

• It is estimated that 40–50% of women with polycystic ovary syndrome (PCOS) are obese [Lobo and Carmina, 2000]. Obesity worsens insulin resistance and increases the risk of Type 2 diabetes and cardiovascular disease. Unfortunately, weight loss is difficult to achieve.
• Weight gain is associated with worsening of symptoms, whereas weight loss may improve the endocrine and metabolic profile [Balen, 2000].
• Limited evidence from small randomized controlled trials and a within-group comparison study suggests that even moderate weight loss can lead to an improvement in hyperinsulinism and hyperandrogenism.
• Good evidence from a Cochrane systematic review and small prospective studies suggests that weight reduction in infertile obese women with PCOS improves ovulation and the chances of pregnancy.
• Expert opinion from a textbook is that if more than 5% of body weight can be lost, there is an excellent chance of restoring menstrual regularity, and an increased incidence of pregnancies has been reported [Balen and Glass, 2005].

Use of orlistat and sibutramine

• Guidelines from the Royal College of Obstetricians and Gynaecologists recommend the use of either of these drugs, as they may reduce body weight and hyperandrogenism in women with PCOS [RCOG, 2007].
• Very limited evidence from a small randomized controlled trial and two small prospective studies suggests that orlistat and sibutramine may reduce body weight and hyperandrogenism in women with PCOS.

Use of bariatric surgery

• This intervention is suggested in guidelines from the Royal College of Obstetricians and Gynaecologists for selected women with morbid obesity [RCOG, 2007]. A small prospective non-randomized study followed up 12 of an original cohort of 17 women with PCOS and morbid obesity who had undergone bariatric surgery. The mean weight loss was 41 +/– 9 kg, and menstrual cycle regularity was restored in all 12 women. In one woman, diabetes and dyslipidaemia resolved, and in another woman, the diabetes reverted to glucose intolerance. However, one of the women undergoing bariatric surgery died of post-operative complications [Escobar-Morreale et al, 2005].

Encourage weight loss

• This recommendation is based on expert advice in a non-systematic review [Balen and Anderson, 2007].
• If more than 5% of body weight can be lost, there is an excellent chance of restoring menstrual regularity [Balen and Anderson, 2007].

Risk of endometrial cancer

• Severe oligomenorrhoea and amenorrhoea in the presence of premenopausal levels of oestrogen can lead to endometrial hyperplasia, which in some women could develop into endometrial cancer [RCOG, 2007]. Inter-menstrual intervals of more than 3 months may be associated with endometrial hyperplasia [RCOG, 2007].
• The true risk of endometrial cancer in women with polycystic ovary syndrome (PCOS) is unknown. Studies are small or have been done on women with infertility, including those with causes of infertility other than PCOS [Balen, 2001]. Other studies have been uncontrolled, and their interpretation is complicated by the variety of diagnostic criteria used to define the syndrome [Hardiman et al, 2003]. There are anecdotal reports of endometrial cancer developing in teenagers with amenorrhoea [Wild, 2002a].
• Endometrial cancer has a mean age of occurrence of 61 years in the UK. Cases in women younger than 35 years of age are exceptionally rare and usually occur when anovulation is secondary to PCOS or oestrogen-secreting tumours [Balen, 2000].

Induce withdrawal bleeding and then refer for ultrasonography to assess endometrial thickness

• These recommendations are based on expert opinion in a review article [Balen, 2000] and advice from CKS expert reviewers.

Prescribe treatment to induce withdrawal bleeding

• These recommendations are based on guidance from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007] and expert opinion in a review article [Hardiman et al, 2003].
• Although evidence is lacking about the risk of endometrial cancer in women with oligomenorrhoea or amenorrhoea, expert opinion is that there is little option other than to advise women to take treatment to induce regular bleeding, at least every 3 months [RCOG, 2007].
• Another option is a progestogen-secreting intrauterine system, such as the levonorgestrel-releasing intrauterine system [Balen and Glass, 2005].
• Cyclical progestogen treatment using less-androgenic progestogens, such as medroxyprogesterone acetate, can be used to induce withdrawal bleeding [Balen, 2000]. However, evidence is lacking about the optimal progestogen, dosage, or treatment regimen [Hardiman et al, 2003]. Guidelines from the Royal College of Obstetricians and Gynaecologists recommend at least a 12-day course of progestogens each month. Medroxyprogesterone is recommended because it is licensed for endometrial protection from oestrogenic hormone replacement therapy as a 14-day course within each 28-day oestrogen hormone replacement therapy cycle.

Women unwilling to take cyclical hormone treatment

• These recommendations are based on expert advice in a non-systematic review [Balen, 2001].

Women who are amenorrhoeic do not need osteoporosis prophylaxis

• Women with PCOS have some ovarian activity [Crosignani and Vegetti, 1996]. Follicular development and oestrogen production continue but are arrested at some stage short of full maturation of an ovulatory follicle. Therefore, although these women are anovulatory, they do not show signs of oestrogen deficiency [Baird, 1997].
• A small case-controlled study of 45 adolescent women with amenorrhoea or oligomenorrhoea (14 had polycystic ovaries) matched with 45 women with regular menstruation found that, although overall the women with amenorrhoea or oligomenorrhoea had lower bone mineral density than those with regular menstruation, those with polycystic ovaries had a bone mineral density similar to that of the control group [To and Wong, 2005].
• A subgroup of 51 women with PCOS in a case-controlled study found that amenorrhoeic women with PCOS had only a marginal decrease in bone mass. Bone protection was thought to be due to adequate oestrogen production and overproduction of androgenic steroids [Adami et al, 1998].

These recommendation are based on advice in guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007], a non-systematic review [Koulouri and Conway, 2009], a textbook [Balen et al, 2005], and guidelines from the Endocrine Society (US) [Endocrine Society, 2008].

Causes of hirsutism

• The most common underlying causes for hirsutism are polycystic ovary syndrome (PCOS) or idiopathic hirsutism [Koulouri and Conway, 2009].
• Hirsutism is present in 60–80% of women with PCOS [Archer and Chang, 2004].
• Other causes of hirsutism are rare and include late-onset congenital adrenal hyperplasia, androgen-secreting tumours (ovarian or adrenal), Cushing's syndrome, acromegaly, or drugs.

Test for elevated androgens

• This recommendation is based on guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007].

Advise weight loss if the woman is obese

• Good evidence from a systematic review suggests that obesity negatively influences treatment for hirsutism.

Cosmetic measures

• Cosmetic measures may be helpful because drug treatments may take 6–9 months or longer before any improvement is noticed [Balen et al, 2005].
• Limited evidence from a Cochrane systematic review suggests that some laser and photoepilation treatments may lead to temporary short-term hair removal.

Treatment of hirsutism

• As hirsutism is caused by excess circulating androgenic hormones, it can be treated with anti-androgenic agents. Other topical and systemic drugs can also be used to treat the specific symptoms. The principal treatment for hirsutism caused by PCOS is to reduce or block the effect of circulating androgenic hormones.

Choice of hormonal treatment

• Oestrogen and progestins suppress gonadotrophin secretion from the pituitary, resulting in decreased androgenic hormones. In addition, oestrogen increases sex hormone-binding globulin levels, which results in decreased free testosterone levels. Progestins can act as an androgen antagonist.
• Combined oral contraceptives (COCs) may be used to reduce ovarian androgen production and treat hyperandrogenism. Some COCs may be more effective than others because of their progestogen content (although there is little direct evidence to support this idea) [Archer and Chang, 2004]. No specific COC has been shown to be superior in treating hirsutism in PCOS, and the best COC for women with PCOS is unknown [Balen et al, 2005].
• The rationale for using COCs to treat the symptoms of PCOS remains controversial. Whilst treatment is undoubtedly of benefit for many symptoms, there is some concern that use of COCs may negatively affect insulin resistance, glucose tolerance, vascular reactivity, and blood coagulation [Ehrmann, 2005].
• CKS expert reviewers recommend co-cyprindiol (Dianette^®) or a COC containing drospirenone as the preferred COCs for women with hirsutism.
• Co-cyprindiol contains the anti-androgen cyproterone acetate.
• Cyproterone acetate inhibits 5-alpha reductase activity, increases sex hormone-binding globulin levels, and has significant anti-gonadotrophin effects.
• There is limited evidence from a Cochrane systematic review that suggests that co-cyprindiol is more effective than placebo for treating hirsutism.
• Co-cyprindiol is licensed for the treatment of moderately-severe hirsutism [ABPI Medicines Compendium, 2008].
• Drospirenone also has anti-androgenic properties [Martin et al, 2008].
• COCs containing drospirenone (such as Yasmin^®) may be an alternative to co-cyprindiol in women with hirsutism, especially as long-term treatment is often necessary.
• There is conflicting poor quality evidence for its use in women with PCOS.
• CKS expert reviewers did not recommend second generation COCs (containing levonorgestrel and norethisterone) and third generation COCs (containing desogestrel, norgestimate, and gestodene) for the management of hirsutism.
• COCs containing levonorgestrel and norethisterone are more androgenic and could potentially exacerbate hirsutism [Koulouri and Conway, 2009].
• There is some concern that COCs containing desogestrel, norgestimate, and gestodene may have a greater risk of venous thromboembolism than those containing drospirenone, levonorgestrel, or norethisterone, although the absolute risk is is still low (about 25 per 100,000 women per year of use) [BNF 57, 2009].

Treatment of relapse when co-cyprindiol is stopped

• The advice on whether to continue to use co-cyprindiol continuously or intermittently, or to switch to an alternative COC, is advice based on the opinions of CKS expert reviewers.

Onset of improvement with drug treatment

• An individual hair follicle takes months to pass through the anagen (active growing), catagen (involutional), and telogen (resting) phases. All systemic treatments for hirsutism reduce stimulation of the anagen growth phase by testosterone, but enough follicles must enter the anagen phase before a clinical effect is noticeable [Koulouri and Conway, 2009].

Length of hormonal treatment

• Co-cyprindiol (Dianette^®) may take 6 months or longer to produce an improvement in hirsutism, so advise the woman to be patient during this period [Balen et al, 2005; Butts and Driscoll, 2006].
• The Committee on Safety of Medicines has recommended that co-cyprindiol should be discontinued three or four menstrual cycles after the woman's acne or hirsutism has resolved, owing to the risk of serious adverse effects, such as thromboembolism [CSM, 2002].
• There is a two- to four-fold increase in the risk of venous thromboembolism with co-cyprindiol compared with conventional second-generation COCs, although the absolute risk remains low [Vasilakis-Scaramozza and Jick, 2001; Seaman et al, 2003].
• The recommendation to continue to use co-cyprindiol continuously or intermittently, or to switch to an alternative COC, is pragmatic advice from CKS.

Eflornithine

• Eflornithine is an irreversible inhibitor of ornithine decarboxylase, an essential enzyme involved in the production of hair [Moghetti et al, 2000]. Inhibition of this enzyme reduces hair growth.
• Limited evidence from randomized controlled trials suggests that it is effective at reducing hirsutism, although this effect is rapidly reversed after stopping treatment [Moghetti and Toscano, 2006].
• No data are available on the long-term safety of eflornithine [Butts and Driscoll, 2006].
• The advice to consider prescribing topical eflornithine for women in whom standard treatment is ineffective, contraindicated, or considered inappropriate is consistent with the restricted use advice given by the Scottish Medicines Consortium [Scottish Medicines Consortium, 2005].
• Limited evidence from a Cochrane systematic review suggests that some laser and photoepilation treatments may lead to temporary short-term hair removal.

Treatments available in secondary care

• Weak evidence from systematic reviews suggests that spironolactone and other anti-androgens are effective in the treatment of hirsutism.
• Spironolactone has moderate anti-androgenic effects. It is not licensed in the UK for the treatment of hirsutism.

Referral

• These recommendations are based on expert advice in a review article [Koulouri and Conway, 2009].

Use of a hormonal contraceptive

• A Cochrane systematic review provides evidence that combined oral contraceptives (COCs) are effective in the treatment of acne vulgaris. In general, the efficacy of COCs in treating acne vulgaris is independent of the particular oestrogen or progestogen component used; however, evidence suggests that the addition of cyproterone acetate as the progestogen component may further increase effectiveness.
• In most women, the safety profile of COCs is good, and they have additional benefits for women who require contraception.

These recommendations are based on advice from the National Institute for Health and Clinical Excellence for obese women in general, including those with polycystic ovary syndrome (PCOS) [National Collaborating Centre for Women's and Children's Health, 2004].

Weight loss

• Lifestyle modification through sensible eating and exercise may improve insulin sensitivity and restore ovulation, even though minimal weight loss is achieved [Balen et al, 2006; Balen and Anderson, 2007].
• Good evidence from a Cochrane systematic review suggests that weight reduction in infertile obese women with PCOS improves ovulation and the chances of pregnancy.
• Most women with PCOS will ovulate in response to clomifene citrate. An increased body mass index is the only factor which is consistently associated with a decreased response to clomifene citrate. Therefore, weight reduction is also an important part of treatment for anovulatory women who are considering treatment with anti-oestrogens [Kousta et al, 1997].

Clomifene citrate

These recommendations are based on guidelines from the National Institute for Health and Clinical Excellence for people with fertility problems [NICE, 2004] and a review article [Balen and Rutherford, 2007].

• Anti-oestrogens, such as clomifene citrate and tamoxifen, occupy hypothalamic oestrogen receptors but do not activate them. This interferes with the binding of estradiol and thus prevents negative feedback inhibition of follicle-stimulating hormone secretion.
• Randomized controlled trials provide evidence from randomized controlled trials that clomifene citrate is effective at increasing pregnancy rates.
• Women may benefit from receiving clomifene citrate in up to 12 cycles, as cumulative pregnancy rates continue to increase after six treatment cycles before reaching a plateau similar to that of the normal fertile population by cycle 12. Alternative treatments should be considered after 12 cycles if results are poor [National Collaborating Centre for Women's and Children's Health, 2004]. This is because evidence suggests that the use of clomifene citrate for more than 12 cycles is associated with an increased risk of ovarian cancer.

Ovarian hyperstimulation syndrome (OHSS)

• Mainly retrospective and prospective studies provide evidence that women with polycystic ovary syndrome (PCOS) are at increased risk of developing OHSS. This is characterized by massive enlargement of the ovaries and ascites, that can lead to rapid and symptomatic enlargement of the abdomen, intravascular contraction, hypercoagulability, and systemic organ dysfunction [Legro, 2001].
• OHSS has been reported rarely in women who have taken clomifene citrate [Brown et al, 2005].

Laparoscopic ovarian drilling (LOD)

• There is evidence that LOD is less likely than gonadotrophin therapy to result in multiple pregnancies in the management of clomifene-resistant PCOS. However, LOD is effective in less than 50% of women, who then need to be treated with additional ovulation induction.
• LOD is recommended as a second-line option if ovulation induction with clomifene citrate has failed [Thessaloniki ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2008].
• The aim of surgery is to destroy ovarian androgen-producing tissue [Farquhar et al, 2007].
• Intraovarian androgen production decreases, which results in a decreased circulating androgen concentration. A reduction in total and free testosterone by 40–50% after LOD has been found [Pirwany and Tulandi, 2003].
• The destruction of the ovarian stroma seems to have an indirect modulating effect on the pituitary gland. Luteinizing hormone concentrations decrease and follicle-stimulating hormone concentrations tend to increase, thus restoring the normal luteinizing hormone/follicle-stimulating hormone ratio. This leads to recruitment of a new set of follicles and resumption of normal ovarian function [Pirwany and Tulandi, 2003].
• These endocrine changes occur rapidly after surgery and are sustained for several years [Pirwany and Tulandi, 2003]. However, there is ongoing concern about the long-term effects of LOD [Farquhar et al, 2007].
• During LOD, either laparoscopic ovarian cautery or laser vaporization (using CO₂, argon or Nd:YAG lasers) is used to create multiple perforations (about 10 holes per ovary) in the ovarian surface and stroma. The procedure can be done on a day-case basis. There is a risk of peri-adnexal adhesions, and LOD should only be offered as a second-line treatment for clomifene citrate-resistant anovulatory women [Pirwany and Tulandi, 2003; Farquhar et al, 2007].
• Ovarian wedge resection is no longer recommended. This was the first established surgical treatment for PCOS and has been used for years but is now obsolete because of the risk of post-surgical adhesions and the introduction of medical ovulation induction [Farquhar et al, 2007].

Gonadotrophins

• There is evidence that gonadotrophins are effective in inducing ovulation in anovulatory women with PCOS who have not responded to clomifene citrate.
• Gonadotropins require careful monitoring to reduce the risk of multiple pregnancy, and low-dose regimens are now used to prevent overstimulation and multiple pregnancies [Balen and Rutherford, 2007].
• Gonadotrophins are recommended as a second-line option if ovulation induction with clomifene citrate has failed [Thessaloniki ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2008].

Multiple pregnancy

• Women with PCOS are at increased risk of multiple pregnancy with both clomifene citrate and human menopausal gonadotrophin treatment [Balen and Rutherford, 2007].

Metformin

• A Cochrane systematic review found no evidence that metformin improves live birth rates when used alone or in combination with clomifene citrate.
• Expert opinion in a consensus workshop [Thessaloniki ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2008] and a non-systematic review [Balen and Rutherford, 2007] is that metformin is no longer recommended as first-line treatment in infertile women with PCOS. Its use should be restricted to women with glucose intolerance [Thessaloniki ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2008].

Screening for gestational diabetes

• These recommendations are based on guidance from the Royal College of Obstetricians and Gynaecologists. Good evidence from a meta-analysis suggests that women with PCOS have an increased risk of gestational diabetes mellitus.

Pregnancy complications

• Good evidence from a meta-analysis suggests that women with PCOS have an increased risk of pregnancy-induced hypertension, pre-eclampsia, and pre-term birth, and that the baby is at increased risk of neonatal complications.

Spontaneous miscarriage after assisted reproduction

• There is evidence that women with PCOS have a higher rate of spontaneous miscarriage after assisted reproduction.

Use of metformin

• The use of metformin in pregnancy is not recommended in guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2007] until further randomized prospective trials are available to provide adequate evidence of safety and efficacy.