Management of Adolescents with Polycystic Ovary Syndrome

Polycystic ovary syndrome (PCOS) is a disorder of metabolic as well as reproductive function; common (although not universal) comorbidities include insulin resistance, hyperinsulinemia, hypoadiponectinemia, dyslipidemia, hepatic steatosis, carotid intimal medial thickening, impaired glucose tolerance, and a predisposition to type 2 diabetes and cardiovascular disease (1, 2). Studies in monozygotic and dizygotic twins (3) suggest that PCOS emerges from the interaction of genetic and environmental factors; the latter may include intrauterine as well as nutritional determinants (1). High-risk groups include low birth weight girls who develop precocious adrenarche and children with family histories of PCOS (4). Sixty percent of patients with PCOS are overweight or obese, and even “normal-weight” subjects have relative visceral adiposity and heightened production of inflammatory adipocytokines (5, 6). Obesity itself causes insulin resistance and may promote free testosterone excess and oligomenorrhea/anovulation through up-regulation of ovarian and adipose tissue 17-hydroxysteroid dehydrogenase 5 expression and down-regulation of hepatic SHBG (1, 7); thus, it is not entirely clear which, and to what extent, the clinical features of PCOS derive from excess adiposity (1, 7–9). Some studies find that insulin resistance and metabolic dysfunction are more severe in adolescents and women with PCOS than in body mass index (BMI)-matched subjects without hirsutism and menstrual irregularity (1, 9). Yet the hyperandrogenism and metabolic defects of PCOS are clearly aggravated by excess weight gain and visceral fat accumulation (1, 5–9). Thus, control of body weight and white adipose storage remains a primary goal in the management of all patients with PCOS. What is the best way to achieve reductions in total body and visceral fat while countering androgen excess, restoring menstrual function, and preventing or reversing the metabolic comorbidities of PCOS? Studies in adults with PCOS show that loss of 5–10% of body weight through diet and exercise counseling can reduce androgen levels and improve menstrual function in approximately 50% of cases (1, 10). Thus, lifestyle intervention is considered by many to be the first line of treatment for PCOS in adults. However, simple weight loss may fail to normalize insulin secretion, glucose tolerance, or dyslipidemia in adults with PCOS and does not always restore menstrual cyclicity or reduce hirsutism scores. Are lifestyle approaches effective for the treatment of PCOS in adolescence? The study by Lass et al. (11) in this issue of JCEM suggests that a lifestyle program consisting of nutrition education (using a generic “traffic light” diet that focused on reducing caloric density), exercise training, and behavior therapy can provide reproductive and metabolic benefits in some obese teenage girls with PCOS. Among 59 patients who completed the 12-month program, the 26 who lost weight (minimum BMI change, −0.2 sd; mean BMI, −3.9 kg/m2; mean BMI z-score, −0.63) had significant reductions in blood pressure (−9 mm Hg systolic, −8 mm Hg diastolic), plasma insulin (−6 μU/ml), and serum triglycerides (−26 mg%); an 8 mg% rise in high-density lipoprotein levels; and an 18% reduction in carotid intimal medial thickness (CIMT). Testosterone levels fell, SHBG levels rose, and menstrual cycles normalized in 61% of the weight responders. In contrast, there was no recovery of menstrual function and no change in metabolic comorbidities in subjects who failed to lose (or gained) weight. These findings are similar to those of Hoeger et al. (12), who showed that modest weight loss (BMI, −1.1 kg/m2) can reduce free androgen index by 59% and increase SHBG by 122% in teenage girls with PCOS. Thus, the available (although admittedly limited) data suggest that lifestyle training programs designed to reduce adiposity may reduce androgen excess, restore menstrual cyclicity, improve insulin sensitivity, reverse or attenuate hyperlipidemia, and reduce blood pressure and CIMT in a subset of girls with PCOS. But the limitations of both studies bear notice. Lifestyle intervention programs are by nature time consuming and personnel heavy. In the Lass et al. study (11), for example, each child attended six group conferences (totaling 9 h) for nutrition education, monthly individual psychotherapy sessions, and weekly play and instructional meetings focusing on the role of physical exercise in daily life. This required an interdisciplinary team of pediatricians, dieticians, psychologists, and exercise physiologists. Yet despite considerable effort, only 59 of the original 127 enrollees completed the yearlong intervention, and only 44% of those who completed the program lost weight. Similarly, the dropout rate was 27% (three of 11 subjects) in the 6-month Hoeger et al. study (12), and an additional 36% failed to attend at least half of the clinic visits. Why are the overall success rates of lifestyle intervention so low? Time requirements, motivational issues, and cultural and socioeconomic determinants no doubt play important roles. But one finding that emerged from the Lass study might be used to advantage: the rate of success of lifestyle intervention varied inversely with the age of the child (11). This observation, common to nearly all studies of lifestyle modification in obese children, suggests that early intervention may prove most effective in reversing maladaptive behavioral trends and preventing irreversible entrainment of metabolic and reproductive dysfunction. The objectives of intervention are prevention or reversal of hyperandrogenism and menstrual dysfunction and reductions in metabolic and cardiovascular risk. Interestingly, the study by Lass et al. (11) suggests that metabolic factors may have differential effects on reproductive and cardiovascular risk in teenagers with PCOS. Levels of free testosterone and SHBG correlated most strongly with plasma insulin concentrations, whereas CIMT correlated far better with diastolic blood pressure and BMI z-score than with serum androgen levels. Likewise, the major determinants of coronary calcium in adults with PCOS (13) were BMI, waist circumference, and total and low-density lipoprotein cholesterol levels, whereas CIMT in adults was variably associated with BMI, cholesterol, or serum androgen levels (14, 15). These observations suggest that weight loss in PCOS may reduce testosterone levels and increase SHBG by increasing insulin sensitivity and reducing circulating insulin levels. In contrast, weight loss may reduce CIMT and cardiovascular risk through reductions in blood pressure and plasma lipid levels. These distinct but overlapping mechanisms must be addressed in any long-term approach to PCOS therapy. What should be done when PCOS develops in nonobese teens or when lifestyle intervention proves ineffective in obese patients? Traditionally, teenage girls with PCOS have been treated with combined estrogen/progesterone oral contraceptives (OCP) and antiandrogens such as spironolactone or (low-dose) flutamide, which increase SHBG concentrations, reduce free testosterone levels, inhibit androgen-dependent hair growth, and restore menstrual cyclicity. OCP may also decrease the long-term risks of endometrial and ovarian cancer but fail to correct the metabolic dysfunction of PCOS in the majority of cases: cholesterol, triglyceride, and C-reactive protein levels may rise (16) and adiponectin levels may fall (17), with persistence (or exacerbation) of severe insulin resistance. OCP can induce weight gain and increase blood pressure in some women and raise the risk of thromboembolism in smokers and those with mutations in Factor V Leiden or high levels of lipoprotein(a). Thus, OCP provide reproductive benefit but confer metabolic risk. The metabolic dysfunction of PCOS may respond to insulin sensitizers such as metformin and the thiazolidinediones, which enhance hepatic and skeletal muscle insulin sensitivity and can reduce circulating insulin levels, increase SHBG concentrations, reduce C-reactive protein, and increase high-density lipoprotein levels (17, 18). As monotherapy, metformin is less effective than OCP in reversing hirsutism and restoring menstrual cyclicity; in combination with OCP, metformin may limit weight gain, reduce fasting insulin and glucose levels, improve glucose tolerance, and attenuate the estrogen-dependent rise in plasma triglycerides. This makes the drug particularly attractive in obese, insulin-resistant teens with PCOS. Nevertheless, limited evidence suggests that metformin and low-dose flutamide in combination may also reduce fat mass, increase insulin sensitivity, and reduce free androgen index in nonobese girls with precocious adrenarche and might delay or prevent their development of PCOS (17, 19). It is currently unclear whether this combination of medications would be effective in other populations of girls or women with PCOS. Likewise, the benefits and risks of glucagon-like peptide-1 agonists in teenagers with PCOS have not been studied, but the combination of exenatide and metformin was more effective than either drug alone in reducing free androgen levels and increasing menstrual cyclicity in adults with PCOS (20). Ultimately, the management of PCOS requires that therapeutic principles be tailored to individual risks and needs. Teenage girls with PCOS should receive intensive lifestyle counseling that emphasizes regular exercise and reductions in the intake of high glycemic foods (which increase insulin and triglyceride levels) and saturated fats (which exacerbate hypercholesterolemia and raise cardiovascular risk). Early intervention is most likely to provide long-term success. Those with longstanding oligo/amenorrhea and/or moderate or severe hirsutism will in most cases require combination estrogen/progesterone and antiandrogen treatment unless there is a personal risk [e.g. Factor V Leiden, increased lipoprotein(a)] or family history of thromboembolic or early cardiovascular disease. Potential indications for metformin include: concurrent obesity that fails to respond to lifestyle counseling, prediabetes (impaired fasting glucose and/or impaired glucose tolerance), and overt type 2 diabetes or a strong family history of the disease. Metformin should also be considered in those in whom OCP are contraindicated. Use of antiandrogens in combination with metformin demands some form of contraception because spironolactone and flutamide are teratogenic. Gastrointestinal complaints in metformin-treated patients can be managed by reducing the dose of medication, although this may limit the drug's benefits; a daily multivitamin is advisable to prevent B vitamin deficiency. Nausea and the need for injections will likely strain the tolerance for glucagon-like peptide-1 agonists in teenagers with PCOS, and there remain lingering (though as yet unproven) concerns about the risk of pancreatitis with this class of medications. Given their risks for weight gain, bone demineralization, anemia, edema, and long-term cardiovascular complications, the systematic use of thiazolidinediones in teenagers must at this time be considered experimental. Finally, and importantly, cardiovascular protection requires intensive control of hypertension and dyslipidemia in all patients with obesity and/or PCOS should these comorbidities fail to respond to other interventions. Body mass index carotid intimal medial thickness oral contraceptive polycystic ovary syndrome. Disclosure Summary: The author has nothing to declare.