Approach

Depending on the needs or goals of the woman, therapy is typically focused on either improvement of fertility or treatment of the symptoms of hyper-androgenism (e.g., hirsutism). Long-term measures should also be taken to restore regular menses and prevent endometrial hyperplasia.

Lifestyle management including a healthy diet and regular physical activity is recommended for all women with polycystic ovary syndrome (PCOS) to optimise general health, quality of life, metabolic health, fertility outcomes, and health during pregnancy.[53][64][78]​​​​[79]​​

With infertility where fertility is desired

The first-line and safest measure to restore ovulation is weight loss (in overweight or obese women). Weight loss alone (even as little as 5% to 7%) may restore ovulation in up to 80% of overweight or obese women (possibly by reducing hyperinsulinaemia and thus hyper-androgenism).​[64][80]​​​​[81][82][83]​​​​​​ Weight loss is also beneficial from a cardiovascular standpoint, and may improve subsequent pregnancy outcomes.[84] Studies suggest dietary interventions, exercise, and/or behavioural coaching are effective for weight loss in PCOS, but no particular exercise or dietary composition (beyond caloric restriction) can be recommended over another.[64][78][85]​​

If weight loss is unsuccessful, pharmacological ovulation induction therapy is recommended. The International PCOS Network and American College of Obstetricians and Gynecologists guidelines recommend letrozole as the first-line option.[53][64]​​ Increasing data suggest that letrozole improves ovulation, pregnancy, and live birth rates compared with clomifene.[86][87][88]​​ However, the use of letrozole may be off-label in some countries, and some guidelines recommend clomifene as the preferred option.[89]

Alternative first-line options include clomifene plus metformin (preferred to clomifene alone), clomifene alone, or metformin alone (less effective but low cost and no monitoring).[53] Gonadotrophins are primarily recommended as a second-line option if other pharmacological treatments are ineffective, but the 2023 International PCOS Network guideline recommends that they may be considered first line as an alternative to clomifene with or without metformin, acknowledging the increased cost, expertise, and monitoring requirements, and the potential for multiple pregnancy associated with gonadotrophin treatment.[53]

Guidelines recommend optimising pre-conception health and lifestyle for all women with PCOS, but weight loss is not recommended as first-line fertility treatment for normal-weight women with PCOS.[53] In these women, letrozole or clomifene should be first-line.

Letrozole

Aromatase inhibitors such as letrozole reduce the conversion of androgens to oestrogens. This reduction in oestrogen synthesis reduces oestrogen negative feedback on the hypothalamus/pituitary, allowing follicle-stimulating hormone to increase and stimulate follicle growth and ovulation.[90]​ The Pregnancy in Polycystic Ovary Syndrome II trial (PPCOS II, sample size 750) found that letrozole was superior to clomifene in the live birth rate.[91]​ Meta-analyses of randomised controlled trials have found letrozole to be superior to clomifene for pregnancy, live birth, and ovulation and similar to laparoscopic ovarian drilling for live birth.[86][87][88] [ Cochrane Clinical Answers logo ] ​​​​​​ Rates of miscarriage, ovarian hyperstimulation syndrome, and multiple pregnancies are similar between letrozole and clomifene.[87] [ Cochrane Clinical Answers logo ]

It is hoped that aromatase inhibitors may be effective in women who are resistant to clomifene; however, too few studies comparing letrozole with placebo in such women have been performed to definitively answer this question.[92]​ By contrast, meta-analysis of randomised controlled trials in clomifene-resistant women with PCOS found that letrozole and laparoscopic ovarian drilling were similarly effective in terms of ovulation, pregnancy, and live birth.[93]​ A randomised controlled trial in clomifene-resistant PCOS found that letrozole and the combination of clomifene plus metformin produced similar rates of ovulation and pregnancy.[94]​ One meta-analysis of individual participant data from six randomised controlled trials found that letrozole was superior to clomifene regarding time to pregnancy, rate of pregnancy, and live birth rate, with no interaction between treatment and BMI on the primary outcome of live birth, suggesting letrozole is superior regardless of BMI.[86]

Metformin

Metformin can restore ovulation/menses to the point where conception is possible. However, 6-9 months may be needed for the full effect. Some data suggest that metformin may be less effective in women with body mass index (BMI) greater than 27-32 kg/m².[95][96][97]​ Patient characteristics that may predict metformin response have not been firmly identified. Some experts believe all women with PCOS may benefit, while others would give metformin only to women who are overweight/obese or who appear to have insulin resistance. One meta-analysis found that the rate of clinical pregnancy was slightly higher with metformin treatment compared with placebo (47.7% vs. 42.9%) for non-obese women with PCOS, but concluded more data is needed before metformin is recommended for non-obese women.[98] Metformin appears to increase ovulation and pregnancy rates, but it has not conclusively been found to improve live birth rates.[97] [ Cochrane Clinical Answers logo ]

Meta-analyses suggest that continuing metformin throughout pregnancy may decrease rates of early pregnancy loss and premature birth, with no effect on gestational diabetes, pre-eclampsia, or fetal abnormalities.[99][100][101] However, in a series of randomised trials, children born to women with PCOS who were treated with metformin (from late first trimester to delivery) had increased BMI compared with children born to women with PCOS in the placebo group.[102]

Clomifene

Clomifene is a non-steroidal anti-oestrogen that inhibits oestrogen negative feedback on the hypothalamus/pituitary, which in turn leads to an increase in follicle-stimulating hormone secretion that may allow follicular maturation and ovulation.

Clomifene is a very commonly used fertility treatment and effective in achieving pregnancy.[103] Up to 25% of patients will have clomifene resistance due to ovarian unresponsiveness. There is a 5% to 10% risk of multiple pregnancy. In a clinical trial comparing clomifene, metformin (and clomifene plus metformin), multiple birth occurred in 6% of the clomifene group and 0% of the metformin group (and in 3.1% in the clomifene plus metformin group).[104]

A meta-analysis found that compared with early follicular phase administration of clomifene, administration during the late luteal phase resulted in a higher total number of follicles, yet rates of ovulation and pregnancy were similar.[105]

Dexamethasone may be added to clomifene if adrenal androgen excess is present.

Clomifene plus metformin

If three treatment cycles of clomifene have failed, it is reasonable to add metformin. Some studies, but not all, suggest that adding metformin to clomifene may be efficacious if clomifene alone is unsuccessful. It is also reasonable to start with clomifene plus metformin rather than either agent alone for treatment of anovulatory infertility.[53]

A Cochrane review concluded that clomifene plus metformin results in a 60% higher pregnancy rate compared with clomifene alone, but data for live birth rates are inconclusive.[97] [ Cochrane Clinical Answers logo ]  Another meta-analysis comparing clomifene plus metformin to clomifene alone found the combination yielded a 28% higher clinical pregnancy rate but no differences in live birth rate.[106] However, in two randomised trials, clomifene was similar in pregnancy or live birth rate to clomifene plus metformin.[104][107] In one of these trials, metformin did not affect the dose of clomifene needed to achieve ovulation.[108] In the other trial, subgroup analysis found metformin efficacious for pregnancy in older (age >28 years) women or those with increased central obesity.[109]

While adding metformin to clomifene seems to improve ovulation rates, the impact on live birth rate has been questioned. Other meta-analyses found clomifene plus metformin to increase pregnancy and live birth versus clomifene alone in clomifene-resistant women.[110][111]

Second-line treatment

If these measures fail, injectable treatments such as gonadotrophins should be given. Gonadotrophins (human menopausal gonadotrophins [hMG]: luteinising hormone [LH] plus follicle-stimulating hormone [FSH]) directly act on the ovary, stimulating follicular recruitment and maturation. In women with PCOS who have anovulatory infertility and clomifene resistance, the International PCOS Network recommends that gonadotrophins are preferable to clomifene plus metformin, gonadotrophins alone are preferred to gonadotrophins plus clomifene, and either gonadotrophins or laparoscopic ovarian surgery can be offered.[53]

Treatment with gonadotrophins is associated with a high risk of multiple pregnancies (twins in 20% to 30%, triplets in 1% to 2%) and ovarian hyperstimulation syndrome (OHSS), especially if many follicles reach intermediate size or if serum estradiol is too high. Mild OHSS (abdominal distention, nausea, vomiting, diarrhoea) is common. Severe OHSS may cause extreme cystic ovarian enlargement (pain, haemorrhagic cysts, torsion), vascular hyperpermeability (ascites, hydrothorax, hypoproteinaemia, electrolyte disturbance, hemoconcentration, oliguria, pulmonary oedema), and, in the most severe cases, thrombosis (sometimes at unusual sites, e.g., subclavian or internal jugular vein) or thromboembolism.

Close follow-up and careful dosing are required to avoid OHSS.

In PCOS, lower doses of hMG are used because of the increased risk of OHSS compared with women without PCOS. FSH alone and hMG have similar rates of OHSS, pregnancy, and live birth.[112]​ Polycystic ovarian morphology is a risk factor for OHSS. Therefore, ultrasound evaluation of the ovaries may assist in selecting the initial dose of gonadotrophins.

The step-up and step-down approaches with FSH were compared in clomifene-resistant women with PCOS. The pregnancy rates did not differ, but the step-up approach had higher rates of ovulation and lower rates of OHSS.[113]​ Another trial found a sequential step-up and step-down protocol to have higher pregnancy and lower miscarriage rates then either step-up or step-down protocols.[114]

Gonadotrophins are usually given as sole therapy; however, adding metformin might reduce the risk of ovarian hyperstimulation syndrome.[115][116]​​ Preliminary evidence suggests that taking metformin during ovulation induction with gonadotrophin, followed by timed intercourse or intrauterine insemination, might increase rates of pregnancy and live birth.[116][117][118]​​ [ Cochrane Clinical Answers logo ]

Third-line treatment

In the most difficult cases, in vitro fertilisation (IVF) or laparoscopic ovarian drilling is performed.

Laparoscopic ovarian drilling (the use of electrocautery or laser to reduce the amount of functional ovarian tissue to reduce androgen production, also reduces inhibin production, allowing FSH to rise and stimulate ovarian aromatase) can restore ovulation and result in pregnancy rates of 25% to 65%. While there is no risk of hyperstimulation or multiple births with ovarian drilling, there is a risk of post-operative adhesion formation (much less than previous ovarian wedge resection techniques) and ovarian atrophy.[119]

One meta-analysis comparing laparoscopic ovarian drilling with medical induction of ovulation (including gonadotrophins, clomifene, letrozole, metformin, and others alone and in combination) in women with anovulatory PCOS who had clomifene resistance found a lower live birth rate with laparoscopic ovarian drilling; when the analysis was restricted to trials with a low risk of bias, the live birth rates were similar.[120]​ Furthermore, in women with anovulatory PCOS who had clomifene resistance, laparoscopic ovarian drilling versus medical induction of ovulation was associated with similar rates of pregnancy and miscarriage but lower rates of multiple pregnancy and ovarian hyperstimulation syndrome.[120] [ Cochrane Clinical Answers logo ] ​ 

There is no conclusive evidence that laparoscopic ovarian drilling leads to diminished ovarian reserve or premature ovarian failure.[121]​ Unilateral and bilateral ovarian drilling may have similar efficacy in clinical pregnancy and live birth rates.[122] Ovarian drilling may be most effective in clomifene-resistant women, with BMI <30 kg/m², and preoperative LH above 10 IU/L.[123][124]

In typical IVF protocols, gonadotrophins are given to promote multifollicular growth so that multiple mature oocytes can be aspirated. Despite more frequent cycle cancellation (failure to retrieve oocytes) and higher rates of miscarriage than in controls, women with PCOS have more oocytes obtained per retrieval and similar pregnancy and live birth rates per cycle.[125][126]​ One meta-analysis found that while not impacting pregnancy or birth rates, metformin administration during IVF and intracytoplasmic sperm injection (ICSI) cycles may reduce the risk of ovarian hyperstimulation syndrome and miscarriage, and improve implantation rates.[127][128] The effect of metformin in reducing the risk of ovarian hyperstimulation during IVF or ICSI cycles with metformin has been confirmed in another meta-analysis; however, this meta-analysis did not find that metformin reduced the incidence of spontaneous abortion.[129]

Not desiring current fertility: hyper-androgenic features alone

Weight loss should be encouraged, but is less efficacious for androgenic symptoms than for therapy of infertility or infrequent/reduced menstrual bleeding. All pharmacological therapies for hirsutism should be trialed at least 6 months before making changes in dose, switching to a new medication, or adding medication.[12][130] 

Oral contraceptive pills (OCP)

In most women with hyper-androgenic symptoms, an oral contraceptive pill (OCP: cyclic oestrogen plus a progestin) is an appropriate choice of initial treatment. OCPs are more effective for acne than for hirsutism.

The Endocrine Society hirsutism clinical practice guidelines and the International PCOS Network guidelines do not recommend one particular OCP over another.[53]​​[130]​ OCP therapy modestly inhibits gonadotrophin secretion, and thus gonadotrophin-sensitive ovarian androgen production, and increases hepatic production of sex hormone-binding globulin (SHBG), which further decreases free testosterone. If free testosterone and SHBG are not normalised after 3 months, the possibility of an androgen-secreting neoplasm should be considered.

Levonorgestrel is the most androgenic progestin and OCPs that contain a progestin with androgenic activity (e.g., levonorgestrel, norethisterone) have often been avoided in practice due to concerns they would be less effective for symptoms like hirsutism. This was not observed in meta-analysis.[131] However, levonorgestrel can have an adverse effect on metabolic biomarkers and therefore it tends to be avoided in women with PCOS.[130] Newer, less-androgenic progestins include desogestrel or norgestimate. Drospirenone is a spironolactone analogue with anti-androgenic and anti-mineralocorticoid properties. Pills with these newer, lower-androgenicity progestins may, however, confer a higher risk of venous thromboembolism than older pills.[132][133][134][135] Given this possible risk of venous thromboembolism with newer pills, some practitioners still prefer to use levonorgestrel- or norethisterone-containing pills.

OCPs should not be used before epiphyseal closure. Contraceptive pills should be avoided or used with caution in women with risk factors (e.g., smoking [especially if aged ≥35 years], history of thromboembolism, or migraine with aura). Other reasons for caution include poorly controlled hypertension, diabetes of long duration (>20 years), and diabetes with vascular complications.[136]

Anti-androgens

The Endocrine Society advises against anti-androgen monotherapy as initial therapy for hirsutism because of its teratogenic potential (unless women are on adequate contraception).[130] For women who are not sexually active, have undergone permanent sterilisation, or are on long-acting reversible contraception, initial therapy with OCP or anti-androgens as monotherapy are both options.[130] If monotherapy is to be used, the decision is tailored to the woman's needs, with a particular focus on adverse effects.

Women with severe hirsutism or contraindications to hormonal contraception may need to be considered for treatment with anti-androgens.[31][130]​​​​ Anti-androgens are androgen receptor blockers (e.g., spironolactone, cyproterone) or 5-alpha-reductase inhibitors (e.g., finasteride). Anti-androgens (especially finasteride) should be avoided in pregnancy due to potential for ambiguous genitalia in male fetus. Flutamide is not recommended because of potential hepatotoxicity. Anti-androgens should be used for at least 6 months before judging efficacy.[130] The maximal effect on hirsutism may take 9-12 months (compared with the effect on acne, which usually responds within 2 months). Acne is more responsive to therapy while alopecia is less responsive. Contraceptive measures are advisable given theoretical teratogenicity.

In many cases, a combination of anti-androgen and oral contraceptive may be needed, particularly for hirsutism or severe acne. The combination has the added benefit of preventing pregnancy, while increasing efficacy by targeting two different processes: androgen production and androgen action. The Endocrine Society recommends monotherapy first line for hirsutism, and if symptoms remain after 6 months to add in an anti-androgen.[130]

Metformin

The Endocrine Society advises against using insulin-lowering drugs for the sole indication of treating hirsutism.[130] Meta-analyses suggest that metformin is associated with decreased testosterone and androstenedione levels and increased SHBG levels, with limited evidence of improvement in hirsutism.[131][137][138] One Cochrane review also found metformin may be less effective in improving hirsutism compared with the OCP in women with PCOS who are overweight.[139] Another meta-analysis of 51 studies concluded that metformin (alone or as adjuvant therapy) may improve acne scores.[140]

Adding metformin might improve results compared with monotherapy or dual therapy.[131][139]​​ Thus, for the specific goal of treating hyper-androgenism, it is best suited as add-on therapy to OCPs, anti-androgens, or OCPs plus anti-androgens. In one meta-analysis, metformin plus spironolactone was more effective for reducing BMI and serum androgen levels than metformin alone, but there was no significant effect on hirsutism score or gonadotrophin levels.[141] The 2023 International PCOS Network guideline suggests that the combination of OCPs and metformin may be most beneficial in high risk metabolic groups, including women with BMI >30 kg/m², risk factors for diabetes, impaired glucose tolerance, or high-risk ethnic groups.[53]

To avoid gastrointestinal adverse effects, metformin should be taken with food and the dose titrated slowly over 4-6 weeks. Extended-release metformin has a slightly lower incidence of gastrointestinal adverse effects. Limited evidence suggests that metformin may promote weight loss, particularly at higher doses (>1500 mg/day) and with longer duration of therapy (>8 weeks).[142]

Long-acting gonadotrophin-releasing hormone (GnRH) analogues

In very severe or refractory ovarian hyper-androgenism, GnRH analogues (e.g., leuprorelin) plus oestrogen yield profound suppression of gonadotrophins and suppress ovarian steroid synthesis.[130]​ GnRH agonists are best combined with oestrogen (OCPs) to increase SHBG and protect bones from resultant hypo-oestrogenaemia (women on GnRH without oestrogen replacement may lose 4% to 8% trabecular bone after 6 months) and avoid severe vasomotor symptoms. With the oestrogen replacement, a progestin must also be given to protect the endometrium.[130]

Mechanical hair removal or topical therapy

At any stage of therapy for hirsutism, mechanical or local hair removal is a useful adjunct to remove hairs that do not respond to medical therapy.[130]

To destroy terminal hair follicles, electrolysis (or laser, which works best with light skin and dark hair) is useful after ≥6 months of hormonal therapy has halted the appearance of new terminal hairs.

Topical eflornithine slows growth of facial hair in 20% to 40% of women by 8 weeks.[143]​ It should be discontinued if no results are noted by 4-6 months.

For androgenetic alopecia, topical minoxidil treatment may be effective but must be used for several months.[144]

With both topical eflornithine and minoxidil, benefit subsides if the agent is discontinued.

Not desiring current fertility: infrequent/reduced menstrual bleeding alone

Anovulatory women with PCOS have chronic oestrogenisation without progesterone exposure, leading to risk of abnormal uterine bleeding, endometrial hyperplasia, and cancer.

Therefore, treatments that induce ovulation (e.g., weight loss or metformin) or provide progesterone exposure (OCPs or a cyclic progestin) should be given to these women.

Weight loss is the preferred treatment for overweight or obese women. The OCP or metformin is used if ineffective, or if weight is normal. If an OCP is not tolerated or desired by the patient, or if there are contraindications to an OCP, cyclic progestin should be given.[12]​ A cyclic progestin is also used in refractory cases. 

Among the older OCPs, ethynodiol/ethinyl estradiol is considered low androgenic and may be useful if a larger dose of oestrogen is needed. This higher-oestrogen-dose pill may be able to induce menses in women with persistent amenorrhoea. Persistent amenorrhoea may indicate endometrial atrophy resulting from hyper-androgenism (thin endometrial width on ultrasound). The risk-benefit ratio must be carefully considered when using higher-dose pills.

The International PCOS Network guidelines do not recommend a particular OCP over another.[53]

Not desiring current fertility: hyper-androgenic features plus infrequent/reduced menstrual bleeding

These women are treated with a combined approach:

  • Preferred treatment is weight loss (if overweight) plus OCPs

  • Oral contraceptive plus an anti-androgen is used if this proves ineffective

  • Long-acting GnRH analogue plus oestrogen is used for refractory cases

  • Metformin may be used adjunctively for weight loss plus oral contraceptives, or for oral contraceptives plus anti-androgen

  • For any stage of treatment, topical or mechanical hair removal may be added.

Use of this content is subject to our disclaimer