KatCarney

Birth control pills and PCOS

The clinical description of polycystic ovary syndrome (PCOS) provided by Dr. Carole Stashwick ("Amenorrhea and acne in the adolescent girl: Is it polycystic ovary syndrome?" October 2000) was excellent. I would like to raise concerns about the recommended traditional treatment with oral contraceptives for PCOS, however. This treatment would make sense if the pathophysiology of PCOS were elevated lutenizing hormone (LH) primarily but recent findings have demonstrated that most women with PCOS initially exhibit various degrees of insulin resistance with compensatory hyperinsulinemia, and that this is what is at the root of the pathogenesis of the syndrome (Nestler JE et al: J Clin Endocrinol Metab 1993;76:273).

As early as 1980, Burghen and coworkers (J Clin Endocrinol Metab 1980;50:116) noticed the association of hyperandrogenism with hyperinsulinism, both basal and in response to a glucose tolerance test. It is interesting that genetic studies have shown that probands with PCOS have family members with insulin resistance or diabetes mellitus type 2 (DM2), suggesting that resistance to insulin in PCOS may be due to a specific genetic defect. Furthermore, in PCOS there is a defect of insulin receptors in muscle of the same magnitude as seen in DM2, resulting in a diminished muscular uptake of glucose. As a consequence, ingestion of sugar is followed by an elevated blood sugar level, which overstimulates [beta] cells of the pancreas and, in turn, increases the compensatory secretion of insulin to facilitate the necessary muscular uptake of glucose.

Patients with PCOS and hyperinsulinism maintain a normal blood glucose level for many years, but their excessive insulin level stimulates cells of the ovarian internal theca. That is how the production of testosterone increases. Many studies have confirmed that, in PCOS, insulin excess is not due to increased androgens, but the other way around. Moreover, suppression of androgens, using gonadotropin-releasing hormone (GnRH) agonists in women with PCOS, does not produce any change in their level of insulin or in their insulin resistance, further strengthening the view that hyperandrogenism in PCOS is the cause, not the consequence, of hyperinsulinism. Conversely, suppression of insulin secretion with diazoxide produces a marked lowering of testosterone, demonstrating that, in these patients, hyperandrogenism is directly induced by an excess of insulin. A possible mechanism of action is insulin lowering of sex hormone binding globulin (SHBG), which allows for the increase in free testosterone.

This knowledge has great therapeutic implications: lowering of insulin level results in weight loss, a lower testosterone level, normalization of the menstrual cycle, and amelioration of acne and hirsutism. For instance, reduction of the insulin level with metformin diminishes the secretion of LH, increases threefold the level of SHBG, and diminishes markedly the hyperandrogenic state of the PCOS. What is more, metformin increases the rate of spontaneous ovulation.

What are the therapeutic implications of all this? Birth control pills, the commonly recommended therapy, probably should not be used long term, as they increase insulin resistance and, therefore, can make matters worse over time.(Godsland IZ et al: J Clin Endocrinol Metab 1992;74:670; Korytkowski MT et al: J Clin Endocrinol Metab 1995;80:3327). One can use other treatments, such as spironolactone and a GnRH agonist, that have no such adverse effect on insulin resistance (Elkins-Hirsh KE et al: Fertil Steril 1993;60:634). The reduction of insulin resistance by weight reduction and pharmacotherapeutic intervention with troglitazone (Ehrmann DF et al: J Clin Endocrinol Metab 1997;82:2108) or metformin (Velazques EM et al: Metabolism 1994;45:647) should eventually become our standard of care.

Tomas Jose Silber, MD, MASS

Washington, D.C.

The author replies: Dr. Silber is correct that a number of investigators believe that insulin resistance may be an important causative factor in PCOS in many, if not all, women. In a study among adult women with PCOS, about 30% had impaired glucose intolerance, and 7% to 8% had type 2 diabetes mellitus (Dunaif A: Endocrinol Metab Clin North Am 1999;28:341). Not all hyperandrogenic women in PCOS families have chronic anovulation and insulin resistance, which suggests to many investigators that there may be variable penetrance of a single gene (such as a "PCOS insulin resistance gene"), or that there may be more than one gene that coders insulin resistance (Dunaif A: Endocrinol Metab Clin North Am 1999;28:341). Particularly in nonobese patients, there exists a significant subset of PCOS patients who have persistent hypersecretion of LH that may be unrelated to insulin resistance (Marshall JC et al: Endocrinol Metab Clin North Am 1999;28:295).

It is not clear how to optimally demonstrate insulin resistance in the adolescent with PCOS, especially in the nonobese patient. Should one order a fasting glucose, a postprandial glucose, a glucose tolerance test, including insulin levels? Certainly obese adolescents, and those with the HAIR-AN syndrome (hyperandrogenism, insulin resistance, and acanthosis nigricans) should be considered at high risk for insulin resistance, screened and monitored for glucose metabolic problems, and managed in consultation with a pediatric endocrinologist.

It may be, as Dr. Silber points out, that long-term treatment for a significant segment of PCOS patients will ultimately focus on glucose metabolism and the management of insulin resistance, rather than on the management of the manifestations of hyperandrogenism that present in adolescence: primarily acne and oligomenorrhea or amenorrhea, and later hirsutism. But there are no long-term studies of PCOS identified in early adolescence and managed by oral contraception or any other treatment modality.

Spironolactone (as an antiandrogen) should be considered when acne is severe or when hirsutism or male-pattern alopecia is present. But spironolactone should not be prescribed without extremely good contraceptive protection, because of concerns that spironolactone taken during pregnancy can result in feminization of the male fetus. An oral contraceptive plus spironolactone is therefore a good treatment choice for these patients.

GnRH agonists are recommended only for patients who fail to respond to an oral contraceptive and spironolactone. Importantly for adolescents, GnRH agonists induce symptoms of menopause: night sweats and mood swings. These medications are also associated with lowered bone mineral density in all patients, including those with PCOS.

Not all investigators believe that metformin is helpful in reducing insulin resistance in PCOS (Ehrmann DA: Endocrinol Metal Clin North Am 1999;28:423). Troglitazone has been demonstrated to improve insulin sensitivity, decrease insulin resistance, and decrease androgen levels in women with PCOS, but it has also been demonstrated to cause severe hepatotoxicity, liver failure, and death. Consequently the drug was withdrawn by its manufacturer by order of the FDA in late 2000. The role of the dose cousin rosiglitazone (Avandia), which can also be hepatotoxic, in the management of patients with PCOS (in all or in only those with demonstrated insulin resistance?) needs much further elucidation. And much more research is needed on the effects of these medications in adolescents (Gordon CM: Pediatr Clin North Am 1999;46:519).

My intent in the article is to bring androgen excess syndromes, especially PCOS, to the attention of the general pediatrician, who has a unique opportunity to identify these common problems early so they can be treated and so that we can promote better adult health among these young people. Oral contraceptives are a first-line treatment for adolescents with androgen excess, and can be managed by the primary care pediatrician or family physician These patients should be screened, to the best of our ability, for insulin resistance and should be referred for additional metabolic management to the appropriate endocrinologist as indicated.

Advances spurred by research over the next 10 to 20 years will certainly inform our understanding of PCOS and our "best practices" treatment. I do not recommend at this time, however, that primary practitioners prescribe the other treatments suggested by Dr. Silber.

Carole A. Stashwick, MD

Lebanon, N.H.

__________________