A 27-Year-Old Woman With a Diagnosis of Polycystic Ovary Syndrome

DR REYNOLDS: Ms R is a 27-year-old woman who was diagnosed as having polycystic ovary syndrome (PCOS) by a previous physician and is now having difficulty with weight gain.

Ms R began menstruating around the age of 13 years; her menses have always been irregular. When she entered college, she was having approximately 9 to 10 menstrual cycles per year; this pattern only began to bother her when she became sexually active and worried whether she was pregnant. Her longest period of amenorrhea was 4 months; she complained of mild facial acne and minimal facial hair growth. At a routine visit for health maintenance at her college health service, a physician suggested that Ms R's history might be consistent with PCOS. The physician ordered a pregnancy test that was negative, along with some blood tests and a pelvic ultrasound. Ms R's serum follicle-stimulating hormone (FSH), luteinizing hormone (LH), prolactin, and thyroid-stimulating hormone (TSH) levels were normal. The pelvic ultrasound showed numerous small ovarian cysts. Ms R was told that she had PCOS; follow-up testing of her blood glucose and glycated hemoglobin levels were within normal limits. She began taking oral contraceptives, and her cycles became regular.

Ms R transferred her care to our medical center several years ago. She now reports weight gain despite a very active lifestyle, regular exercise, and a nutritious diet. Her weight gain is troublesome to her and she has been unable to keep weight off despite her strict adherence to diet and exercise recommendations.

Ms R has no other medical problems; her only medication is an oral contraceptive. She works as an administrator and project manager, is sexually active with 1 male partner, and does not smoke or use recreational drugs or alcohol. She bikes 20 miles every other day and swims regularly.

Ms R's father developed diabetes in his 60s; her mother has a history of deep venous thrombosis (DVT).

On physical examination, Ms R is 59 inches tall and weighs 122 lb (54.9 kg) (12 lb [5.4 kg] more than at her visit 2 years ago). Her body mass index (BMI) is 25. Her blood pressure is 110/80 mm Hg. Her skin shows very mild facial acne; she does not have hirsutism. Ms R's physician was not able to palpate ovaries at her most recent pelvic examination.

Laboratory testing showed normal levels of free testosterone and dehydroepiandrosterone sulfate (DHEAS), drawn while Ms R continues to take contraceptives. Ms R's serum levels of glucose and her percent glycated hemoglobin were within normal limits. Her fasting total cholesterol value was 167 mg/dL (4.32 mmol/L); her high-density lipoprotein cholesterol (HDL-C) level was 72 mg/dL (1.86 mmol/L). A repeat transvaginal ultrasound showed normal-sized ovaries with multiple peripheral follicles.

Well, I was in college and noticed that I had some menstrual cycles that were really, really variant. Sometimes there was 3 weeks between cycles; other times there were 2; other times it would be 8 or 9 weeks. I do have facial hair growth, but nothing that is terribly excessive, more so than I would consider normal but not by a large amount.

I’ve always done a lot of just wellness visits with my physicians over the years, and I will often go in just for regular checkups, whether I have a particular medical condition or not, and the irregular periods were something that came up frequently during my visits with my physician at the time, who was great. She began to notice a pattern that she was mildly concerned about and wanted to just explore it a little more. No one was really certain what caused all of these different, wide irregularities in my menstrual cycle. My doctor at the time did a lot of testing for various hormone levels that all appeared to be normal. Then eventually, she sent me over to get an ultrasound done to rule out PCOS. Every other test had ruled out the possibility of PCOS, but the ultrasound indicated that there was some cyst formation. So, that was sort of a default diagnosis we arrived at.

She actually didn't talk a lot about PCOS with me, particularly related to fertility. I do want to have children at some stage in the future; I don't want to have children now. At the time we were discussing whether or not I should go on birth control pills, and the PCOS was yet another reason to go onto birth control pills. I’ve always been very averse to taking medications for anything, and so I resisted the idea for awhile, but it was really sort of the PCOS diagnosis that pushed me over into starting to take birth control medication. I know there is a chance that my fertility might be affected by PCOS, but I don't really know that's what will happen. From all the discussions I have had and all the reading I have done, the fertility problem doesn't sound as if it is very conclusive at this point in time.

My weight's been relatively stable, but this year I have noticed that my weight started to creep up, and that has been concerning, because I am incredibly active. I bike ride every day for miles, and I play soccer several times a week. There hasn't been a good reason for my weight to be going up. My diet hasn't changed. So, that is something that my doctor and I have been watching, and we have been sort of watching my blood glucose, which seems to be fine, because there is the concern of diabetes. We’ve been thinking about looking at metformin as a possible sort of sugar regulator to see if that’ll help with the weight, but I haven't started taking that yet. I am averse to taking drugs.

So, my current dilemma with continuing on birth control pills is I’ve become recently more acutely aware of the risk of blood clots. I’ve had a couple of friends who’ve been diagnosed with blood clots. My mother has had blood clotting, and my doctor and I have been discussing whether to stay on birth control pills, which, as I understand it, preserves ovarian function or the reproduction function. That decision is colored by the fact that I’ve been diagnosed with PCOS; I might more readily come off the pills if I weren't aware of that condition.

Does Ms R have PCOS? What is PCOS and what causes it? In which patients is evaluation for PCOS indicated? What tests should be performed as part of the evaluation? What tests affect clinical management? What are the implications of a diagnosis of PCOS for Ms R? Is treatment effective, and in what ways is it effective? Is Ms R's weight gain related to PCOS? What should Ms R understand about her future fertility? What do you suggest for Ms R?

DR LEGRO: Polycystic ovary syndrome is a common endocrinopathy (best described as unexplained hyperandrogenic chronic anovulation) that affects 7% of the reproductive-aged female population (based on a population-based survey of women applying for employment at an academic medical center in Alabama).¹ Because its etiology and natural history are poorly understood, diagnosis and treatment are controversial. Its origins as a named disorder track back to its original description in the 1930s by Stein and Leventhal, who described a syndrome of signs and symptoms including oligomenorrhea, enlarged polycystic ovaries, hirsutism, and obesity.^{2 - 3} Varying definitions of the syndrome and its heterogeneity confound the interpretation of the current literature.⁴

Ms R has many of the features of PCOS including a history of chronic anovulation, androgen excess (based on her complaint of mild acne and hirsutism), and polycystic ovaries on ultrasound. She is taking oral contraceptives, which ameliorate many of the symptoms, but she has concerns about their adverse effects and is searching for other treatments.

Diagnostic criteria, sometimes called the Rotterdam criteria, were developed by an expert panel^{5 - 6} (Table 1) and modify the prior 1990 National Institutes of Health (NIH) criteria⁷ by incorporating ovarian morphology and size into the schema. Androgen excess presents with clinical signs (most commonly by disorders of the pilosebaceous unit, such as acne, hirsutism, or androgenic alopecia) or by biochemical measures (most commonly, elevated circulating testosterone levels). Based on expert opinion, the criteria have not been validated, leading to concerns about some of the phenotypic categories, especially those that exclude hyperandrogenism.⁸ The more liberal Rotterdam criteria expand the pool of potential women with PCOS by 50%.⁹ Individual components of the syndrome have varying reproducibility. Testosterone assays are not standardized and have wide interassay variation, especially when compared with the gold standard measure by mass spectrometry.¹⁰ Ultrasound-based antral follicle counts, which are used to identify polycystic ovaries, have been found to have adequate intraobserver and interobserver reproducibility, although they may be better for categorical assignment.¹¹ Interobserver and intraobserver correlation for assessing hirsutism are poor.¹²

Ms R has a history of mild acne and mild hirsutism, suggesting hyperandrogenism. Clinical hyperandrogenism is dependent on a number of factors, including the observer¹² and race (eg, Asian population has little midline body hair).⁵ Many clinical trials have opted for a more objective measure of biochemical hyperandrogenism.^{13 - 14} Androgen measurement can be problematic in clinical practice where most testosterone assays are insensitive in the range of women with PCOS (60-150 ng/dL [2.08-5.20 nmol/L]).^{15 - 16} Androgen levels are further confounded by age, reproductive maturation,¹⁷ and medications, especially oral contraceptives, which normalize levels of circulating androgens in women with PCOS.¹⁸ Ms R is taking oral contraceptives and, as expected, her testosterone level was normal, although she reports hirsutism.

A history of chronic anovulation is usually obtained by noting the number of spontaneous menstrual cycles over the last year. Most clinical studies use some threshold such as 6, 8, or 9 or fewer menses per year.^{13 - 14 ,19 - 20} Ms R has a history of chronic anovulation when she is not taking oral contraceptives, with approximately 9 menses a year, sometimes going up to 4 months without menses, and also a history of frequent light bleeding episodes as often as every 2 weeks, a pattern that represents anovulatory bleeding.

Polycystic ovaries (usually present bilaterally) are characterized on ultrasonography by multiple (10-12) small follicles (3-8 mm in diameter) usually tightly spaced along the periphery of the ovary, increased central stroma, and increased volume (Figure 1).²¹ The name “polycystic ovary” is a misnomer, because there are actually an absence of “cysts,” ie, no large (>20 mm in diameter) dominant follicle or postovulatory corpus luteum cyst due to anovulation. Ms R had an initial pelvic ultrasound that showed “numerous small cysts” and a follow-up ultrasound consistent with polycystic ovaries. The volume of the ovary was not noted, but this may be the most sensitive and specific marker for polycystic ovaries.²²

Both the NIH and Rotterdam criteria consider PCOS a diagnosis of exclusion (Table 1).^{5 - 7} Tumors may present with rapid and frank virilization, but an elevated testosterone level (>200 ng/dL [6.94 nmol/L]) is nonspecific.²³ Thyroid disease is common in women, can be identified with a TSH measurement, and is readily treated. About 20% to 30% of women with PCOS have mildly elevated prolactin levels in the range of 20 to 30 ng/mL, which probably represents hypothalamic pituitary dysfunction associated with PCOS.²⁴ It is worthwhile to measure 17-hydroxyprogesterone as a screening test for late-onset congenital adrenal hyperplasia due to 21-hydroxylase deficiency, although oral contraceptives may affect circulating levels and a full corticotropin stimulation test may be indicated. Ms R has few stigmata of Cushing syndrome, and given the implications of a false-positive screening test (I prefer a 24-hour urine for urinary free cortisol), may not be indicated in this low-risk patient. Women with PCOS often display elevated LH levels.²⁵ The routine measurement of gonadotropins other than to exclude ovarian failure is also probably not indicated, given their inherent variability and the suppressive effect of increasing obesity,²⁶ or in Ms R's case, the use of oral contraceptives.¹⁸

In summary, there is strong presumptive evidence that Ms R has PCOS based on either the NIH or Rotterdam criteria given her history of chronic anovulation, androgen excess (based on her report of mild acne and hirsutism), and polycystic ovaries on ultrasound (Table 1).

The etiology of PCOS remains unknown. The ovary has been implicated as the prime suspect,² as has the hypothalamic-pituitary axis,^{25 ,27} and currently some defect in insulin action²⁸ (Figure 2). There is a vicious feedback loop in which disordered steroid feedback from both the adrenal and the ovary (primarily androgen and weak peripherally aromatized estrogens) leads to inappropriate hypothalamic-pituitary gonadotropin secretion (abnormal pulsatility with excess LH and normal FSH levels).²⁹ This feedback loop may be exacerbated by insulin resistance.^{30 - 31} Women with PCOS tend to display decreased glucose uptake in response to a given level of insulin, and thus as a group are “insulin resistant,”²⁸ although up to half may not have documented insulin resistance by dynamic testing.³² Compensatory hyperinsulinemia may contribute to the vicious cycle by stimulating adrenal and ovarian androgen biosynthesis,^{33 - 34} by lowering levels of circulating binding proteins (primarily sex hormone–binding globulin) that limit androgen bioactivity,³⁵ or by directly altering gonadotropin secretion.³⁶ The mechanism for selective tissue insulin sensitivity, where some tissues are profoundly resistant (skeletal muscle) and others are sensitive (adrenal and ovary) is unknown, but may involve postreceptor signal transduction pathways that are differentially regulated in varying tissues.^{31 ,37}

Polycystic ovary syndrome is important to diagnose for its immediate impact on reproduction and potentially for its long-term health implications. Polycystic ovary syndrome is the leading cause of anovulatory infertility,¹⁴ and women with PCOS may also be at increased risk for pregnancy loss³⁸ and later pregnancy complications.^{39 - 40} A recent meta-analysis of 720 women with PCOS and 1550 control women that controlled for confounders such as differences in BMI showed women with PCOS demonstrated a significantly higher risk of developing gestational diabetes (odds ratio [OR], 2.94; 95% confidence interval [CI], 1.70-5.08], pregnancy-induced hypertension (OR, 3.67; 95% CI, 1.98-6.81), preeclampsia (OR, 3.47; 95% CI, 1.95-6.17), and preterm birth (OR, 1.75; 95% CI, 1.16-2.62).⁴¹ Nonetheless, obesity remains a significant independent risk factor for subfecundity and pregnancy complications.^{42 - 43} I would counsel Ms R that she is at risk for anovulatory infertility after discontinuing oral contraceptives. She is also at risk for gestational diabetes given her PCOS and family history.

Obesity and insulin resistance, frequently but not invariantly found in women with PCOS, have been implicated in the development of the major causes of death in our society, including increased risks for cancer and cardiovascular disease.^{44 - 45} Randomized US multicenter trials have shown women with PCOS to have a BMI in the range of 35 to 40.^{13 - 14} Women with PCOS from other countries tend to be leaner, in the BMI range of 25 to 28 or less.^{46 - 47} The supersizing of PCOS in the United States may be due to reductions in activity or differences in diet, including composition of diet.⁴⁸ Obesity further exacerbates metabolic and reproductive abnormalities in women with PCOS and may bring out the PCOS phenotype in a susceptible population as suggested by family studies.⁴⁹ Ms R is not obese and displayed the stigmata of PCOS as a normal weight individual, and she has no known familial component other than the history of diabetes.

About 30% to 40% of women with PCOS display glucose intolerance (2-hour glucose level ≥140 mg/dL [7.77 mmol/L]) with oral glucose tolerance testing (2- to 3-fold increased risk compared with weight-matched controls).^{50 - 52} Risk factors for glucose intolerance include a family history of diabetes, age, obesity, and especially a centripetal fat distribution.^{50 - 52} Most women with PCOS and impaired glucose tolerance have normal fasting glucose levels, and the yield is also high in nonobese women (10%),⁵⁰ so this test should be obtained in even nonobese women with PCOS such as Ms R. Glucose intolerance is a much stronger predictor of sudden cardiovascular events in women than in men and is important to recognize and treat.⁵³

Women with PCOS often have an adverse cardiovascular risk profile, including obesity, a centripetal fat distribution, glycemic abnormalities, and dyslipidemia (including low HDL-C, high low-density lipoprotein cholesterol [LDL-C], and high triglyceride levels).⁵⁴ Abnormal clotting and inflammation have been associated with insulin resistance in women with PCOS.^{55 - 57} Studies have also noted an increased prevalence of subclinical markers of atherosclerosis such as increased carotid intimal-medial thickness or coronary calcification.^{58 - 60} However, the evidence for increased or premature cardiovascular events is inferential.^{54 ,61}

The risk profile for endometrial cancer overlaps with PCOS: obesity, especially with a centripetal distribution, diabetes, nulliparity, and a history of chronic anovulation with unopposed estrogen exposure.⁶² However, aside from small case series, little epidemiological evidence links PCOS with endometrial cancer.⁶³ Regardless, Ms R currently has a favorable risk profile for these diseases. She is approaching but not yet overweight, not hypertensive, has normal random glucose and glycohemoglobin values, and also a random lipid profile that was completely normal, with an HDL-C level well above 50 mg/dL (1.29 mmol/L).

There is little evidence that primary treatment of women with PCOS prevents long-term medical complications. Few well-designed, adequately powered trials exist to guide treatment for women with PCOS (Table 2). Much of the impetus for treatment is extrapolated from larger studies in comparative populations. A good example of this is the use of treatments from diabetes prevention trials, including metformin, thiazolidinediones, and lifestyle interventions for women with PCOS.^{70 - 72} Accordingly, treatment should be individualized to the woman with PCOS, and this discussion will therefore focus on Ms R's unique profile.

Ms R is worried about her weight. Weight loss should be a cornerstone therapy for any overweight/obese population at risk for diabetes and cardiovascular disease, but Ms R should be reassured that her weight is not currently a medical issue. Because many women with PCOS are often overweight and experience difficulty losing weight as has Ms R, there is an impression that these are innate characteristics of PCOS. There is little published evidence to show that women with PCOS have difficulty losing weight with dietary therapy. Randomized trials of varying hypocaloric diets in women with PCOS have resulted in appropriate short-term weight loss after 4 weeks (with a 1000-kcal deficit per day designed to lose 1 kg per week, mean weight loss at 4 weeks was 4 kg)⁷³ and after 12 weeks (with a ≈1400 kcal/d diet designed to lose about 0.45 kg per week, participants lost approximately 8 kg),⁷⁴ or with hypocaloric meal replacements followed by a maintenance diet (≈5 kg at 6 months).⁷⁵

These studies did not find a benefit to a particular dietary composition (high protein vs low protein), although their sample size was small and both had high dropout rates (26% in the 4-week trial and 38% in the 16-week trial).^{73 - 74} In an obese population, high-protein diets offer a better short-term (12 week) weight loss, but these differences disappear over longer terms (12 months).^{76 - 77} Few well-controlled, long-term dietary studies have been conducted in women with PCOS. In one study that restricted women to a hypocaloric diet of 1000 to 1500 kcal/d, about half of women lost more than 5% of their starting weight over 6 to 7 months.⁷⁸ Further, neither baseline insulin resistance nor hyperinsulinemia identifies obese women likely to respond to a dietary intervention with weight loss.⁷⁹

The question of whether metformin alone or combined with lifestyle changes enhances weight loss in women with PCOS is tantalizing; like many young women with PCOS, Ms R wonders whether metformin may benefit her. Metformin is a biguanide that decreases hepatic glucose production and also has some peripheral insulin-sensitizing action. Participants in the metformin group of the Diabetes Prevention Program lost about 2 kg over the first 6 months of the trial,⁷⁰ although a meta-analysis of all metformin studies and in PCOS both showed no net weight loss.^{80 - 81} Metformin is not approved by the Food and Drug Administration for the treatment of obesity or for PCOS, and no adequate dose-ranging studies have been performed in women with PCOS. The risks of metformin include gastrointestinal adverse effects such as anorexia and diarrhea, which are reduced with an alternative extended-release preparation.^{82 - 83} Rare events such as lactic acidosis in the presence of renal insufficiency are unlikely to occur in otherwise healthy women with PCOS.^{14 ,84} Few placebo-controlled trials of metformin and lifestyle have been conducted in women with PCOS, but the combined treatment groups have either shown or trended toward a greater weight loss than lifestyle and placebo.^{69 ,85} Therefore, metformin may modestly contribute to weight loss in Ms R.^{85 - 86}

The question of future fertility remains perhaps the most important one for Ms R. She expresses the desire to both preserve and maintain her fertility, but at this moment is pleased with the contraceptive efficacy of oral contraceptives. Ms R is likely to experience continued oligo-ovulation after discontinuing oral contraceptives, as this treatment is not a “cure” for PCOS, but rather a temporary interruption of the vicious circle (Figure 2).

Therefore, Ms R will most likely require treatment to induce ovulation when she is ready to seek fertility. Treatment of each one of the putative mechanisms (ovarian hyperandrogenism, inappropriate gonadotropin secretion, and hyperinsulinemia) can improve ovulation rates in many women with PCOS (Figure 2). There is no evidence-based schema to guide treatment, but my recommendation would be to start medical treatment with clomiphene citrate. Clomiphene has been the longstanding, first-line agent for ovulation induction in PCOS.^{87 - 88} Clomiphene is a triphenylethylene-derived nonsteroidal agent that is theorized to function at the level of the hypothalamus as an antiestrogen to restore gonadotropin secretion.⁸⁹ Its use is associated with hot flashes, mood changes, and (rarely) changes in vision thought to be due to pituitary gland swelling. From a public health perspective, more concerning is the relatively high rate of multiple pregnancy after conception with clomiphene of 7.8%, including a high order (triplets or more) multiple pregnancy rate of 0.9%.⁹⁰ No trials of adequate size have been conducted to comment on whether risk is higher in women with PCOS. Retrospective case series in normal-weight women with ovulatory dysfunction, like Ms R, have shown high conception rates in clomiphene responders approaching 50% after 3 cycles of treatment and 75% within 9 cycles.⁹¹ A meta-analysis has found that clomiphene use is significantly more likely to result in pregnancy than placebo in anovulatory women (number needed to treat, 5.9; 95% CI, 3.6-16.7).⁸⁸ The use of clomiphene to treat anovulation in women not seeking pregnancy has not been well studied and is not recommended currently. Aromatase inhibitors have also been used to induce ovulation in women with PCOS,⁹² and preliminary trials suggest higher pregnancy rates and potentially lower complication rates of multiple pregnancies; larger randomized trials are under way.⁶⁴

Metformin has been shown in a meta-analysis to have a significant benefit on inducing ovulation in women with PCOS.⁸¹ The mechanisms for this effect include decreased ovarian and adrenal androgen production, sequestration of abnormal steroid levels by increased sex hormone–binding globulin levels, and perhaps most importantly, the consistent lowering of insulin levels (Figure 2). When used as a sole agent, ovulation is achieved in 46% of recipients compared with 24% in placebo groups (number needed to treat, 4.4), comparable with the benefit conferred by clomiphene.⁸¹ Metformin also appears effective at improving ovulation in thinner women and adolescents with PCOS.^{67 ,93} Unfortunately, few head-to-head trials have compared clomiphene with metformin.^{14 ,66}

The use of metformin is not associated with any fetal toxicity or teratogenecity and it is a category B drug. It is unknown, but theorized, that conception after treatment with metformin may have a lower multiple pregnancy rate than with clomiphene citrate.^{94 - 95} Metformin has also been touted to decrease miscarriage rates in women with PCOS when continued throughout early pregnancy, and a recent meta-analysis supports a benefit.⁹⁶ Prospective randomized trials are needed to support its use, given the variation between meta-analyses based on small trials and subsequent larger trials.^{65 ,81} The combination of metformin and clomiphene was initially shown in a meta-analysis to significantly improve pregnancy rates.⁸¹ Subsequent larger randomized controlled trials have shown no benefit to combination therapy over clomiphene in terms of pregnancy rate,⁶⁵ and I would not recommend it as initial therapy nor would I recommend surgical therapies, such as ovarian diathermy, as first-line infertility treatment for Ms R.⁹⁷

Ms R expresses the desire to control her disorder as simply and with as few medications as possible.

I suggest that Ms R continue taking her oral contraceptives. She is satisfied with her regular and predictable withdrawal bleeds and values their contraceptive efficacy. Her hirsutism and acne have remained stable with this treatment. One benefit in continuing oral contraceptives is their profound prevention of endometrial cancer, with a reduction of risk by 56% after 4 years of use and 67% after 8 years in users compared with nonusers.⁹⁸ Ms R desires to lose weight and she may be reassured that use of the pill is not associated with weight gain.⁹⁹ In fact, aging represents one of the greatest risk factors for weight gain.¹⁰⁰ Ms R has a favorable risk factor profile for serious clotting complications with the pill, as she is young, nonobese, does not smoke, and is normotensive.

Some researchers maintain that oral contraceptives have adverse effects on diabetes and cardiovascular risk in women with PCOS.¹⁰¹ It may be prudent to measure a 2-hour glucose level after a fasting 75-g oral glucose challenge, as Ms R is at increased risk for glucose intolerance due to her PCOS and her family history of type 2 diabetes.

Above all, I would counsel Ms R that she appears to have a mild case of PCOS (ie, no severe metabolic sequelae), that her symptoms appear to be well controlled with the oral contraceptives, and she has, comparatively speaking, done an admirable job at weight control: a BMI of 25 is a great place to start a weight loss program. I would refer Ms R to a dietitian for further counseling on diet to achieve her desired weight. I would counsel against excessive “Googling” of PCOS on the Web, express optimism that she would respond favorably to the above recommendations, and reassure her that the outlook for her future fertility, most likely with treatment, is excellent.

QUESTION: Do you have any preferences in terms of birth control pills to treat PCOS?

DR LEGRO: I don't know which birth control pill is best, as there are a lack of head-to-head trials in PCOS (or for other gynecological disorders).¹⁰² Now theoretically, the pill that's being hyped out there for women with hyperandrogenism is Yasmin (a 21-day monophasic pill with 30 μg of ethinyl estradiol/3.0 mg of drospirenone [Berlex Inc, Wayne, NJ] or a newer lower-dose version, Yaz (a 24-day monophasic pill with 20 μg of ethinyl estradiol/3.0 mg of drospirenone [Berlex]), although there are only observational, open-label trials to support their use in adult women with PCOS.¹⁰³ Yasmin contains a progestin, drospirenone, that also has activity as an antimineralocorticoid and as an antiandrogen, and pharmacologically it is related to spironolactone.¹⁰⁴ Since the combination of oral contraceptives and spironolactone is frequently used to treat hirsutism,^{105 - 106} I can argue that the use of a pill containing a progestin that is also an antiandrogen saves the patient from additional medication and expense.

However, I’m always concerned about the risks of a new pill and a new progestin, especially with what we’ve gone through with reports about increased clotting events with the third-generation progestins (desogestrel and gestodene) compared with earlier progestins in older pills.¹⁰⁷ It's too soon to say what the event rate is with Yasmin.¹⁰⁸ There is a published report from the Netherlands that raised concern about increased clotting events.¹⁰⁹ Again, I don't know whether that's a detection bias because of the heightened surveillance with a new formulation or a selection bias involving women with PCOS/hyperandrogenism, who may have an increased risk for clotting events (ie, obesity, hypertension, the metabolic syndrome with abnormal clotting) receiving this pill. Currently, until evidence to the contrary appears, it would be my pill of choice, but the evidence would equally support the use of any oral contraceptive.

QUESTION: You mentioned that you weren't worried about a DVT at this stage because she’d already been on the pill. But had you seen her initially with her family history and her story, would you have considered the pill? And as a follow-up, is there any place for progesterone-only treatment?

DR LEGRO: That's a good follow-up question. Ms R mentioned in the full interview that her mother had been screened after her DVT and there was also no thrombophilia present in other relatives. So to me, Ms R's mother had an isolated instance of DVT and I would feel comfortable putting Ms R on a normal monophasic contraceptive pill. I would choose a progestin-only pill for a patient with a strong family history and, in some cases, a personal history of thrombophilia. But I would not use a progestin-only pill routinely in women with PCOS. One of the reasons for this is because the bleeding pattern is often worse on progestin-only hormonal medications,¹¹⁰ especially with Depo-Provera (medroxyprogesterone acetate injectable suspension).¹¹¹ It is hard to convince a woman with PCOS that she is better off with frequent and unexpected bleeding episodes than she is with complete amenorrhea.

QUESTION: I’m hearing that you don't see any role at the outset for metformin unless there's glucose intolerance?

DR LEGRO: Clinically in a patient not seeking fertility, I use metformin for glucose intolerance because to me it's been shown to prevent diabetes given this preexisting condition.⁷⁰ If Ms R also had impaired glucose tolerance, I could argue that metformin may be the next treatment step for her, as she is someone who is already physically active and careful about her diet.

QUESTION: Is there any utility in measuring insulin levels for patients with PCOS?

DR LEGRO: No, I think the clinical utility is in discovering glucose intolerance. Insulin levels have not been incorporated into the diagnosis and management of the metabolic syndrome due to their variability and lack of a standardized assay.^{32 ,112} Glucose intolerance is common in women with PCOS^{50 - 52} and may present with normal fasting insulin levels due to beta cell dysfunction.^{32 ,113} In women with glucose intolerance, normal insulin levels are thus falsely reassuring. Hyperinsulinemia is also unnecessary for treatment selection. It does not predict response to insulin-sensitizing agents in PCOS; some might argue the opposite.¹³ So I do not routinely check insulin levels or use them in clinical management. I think it's just another surrogate marker without a clear clinical validation in the treatment of PCOS.

QUESTION: Can you comment on the sexual function of women with PCOS?

DR LEGRO: That's a great question and I don't know the evidence-based answer to that, but I suspect that problems are underrecognized. Studies using validated quality of life questionnaires have shown that women with PCOS have a lower self-esteem due to problems with excess body hair, due to obesity, and due to the lack of a normal cycle and associated fertility problems.¹¹⁴ One such study reported normal frequency of intercourse, but decreased sexual satisfaction.¹¹⁵ But there are few detailed studies of their libido and sexual function.