| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Special Articles |
Department of Pediatric Endocrinology Columbia University New York, New York 10032
Address correspondence and requests for reprints to: Sharon E. Oberfield, M.D., Professor of Pediatrics, Department of Pediatric Endocrinology, Columbia University, 630 West 168th Street, PH-5 East Room 522, New York, New York 10032.
 |
Introduction |
|---|
 Top Introduction References |
|---|
10% of reproductive age women, is
characterized by hyperandrogenism and chronic anovulation. It is often
associated with obesity, dyslipidemia, and acanthosis nigricans (AN).
Insulin resistance is a common comorbidity in women with PCOS and is
associated with increased risk for hypertension and cardiovascular
disease. Studies have shown that 25–35% of obese women with PCOS, by
30 yr of age, will have either impaired glucose tolerance or type 2
diabetes (2). It has been suggested that PCOS should be
considered as another metabolic abnormality of "Syndrome X," which
includes dyslipidemia, hypertension, and insulin resistance
(3). The actual role of insulin (hyperinsulinemic insulin resistance) in the development and maintenance of hyperandrogenism in PCOS is a controversial area in which consensus has not been reached. In vitro and in vivo studies have shown that insulin acts at multiple sites, often leading to increased androgen excess and effect. Additionally, in women with PCOS, genetic factors, which may modulate the response to insulin at the cellular level, are just beginning to be described (for reviews see Refs. 4 and 5). An increase in serine/threonine phosphorylation of P450c17 increases the enzyme’s 17,20 lyase activity, resulting in an increase in androgen production, as seen in adrenarche (6). Increased serine phosphorylation of the insulin receptor ß chain causes insulin resistance by inhibiting tyrosine phosphorylation (7). The contemporaneous reports of Miller suggesting and Dunaif demonstrating that some women with PCOS have hyperphosphorylated insulin receptors allowed for speculation of the existence of a common molecular pathway—hyperactivity of a single serine kinase—for the two major features of PCOS, namely hyperandrogenism and insulin resistance (6, 7, 8). Recently, however, this was not confirmed in analysis of 17,20 lyase activity in cultured fibroblasts obtained from women with PCOS (9).
Insulin has been shown to directly stimulate ovarian androgen secretion through effects on steroidogenic enzymes (10). Nestler et al. (11, 12) have shown that these effects were partially reversed with use of metformin. It has been suggested that insulin also augments adrenal androgen synthesis (13). In vitro studies have demonstrated that insulin may also stimulate LH release directly or enhance LH release from pituicytes, thereby indirectly stimulating ovarian androgen production (14). Insulin may increase ovarian LH receptors (15). Insulin is known to decrease hepatic sex hormone-binding globulin (SHBG) production, allowing for an increase in the levels of circulating free androgens (16). Additionally, insulin has been shown in vitro to decrease insulin-like growth factor-binding protein I (IGFBP-I), which results in an increase in IGF-I, which may then act locally at the ovary to stimulate androgen synthesis (17). Thus, reduction of serum insulin would be expected to decrease some of the androgen-related symptomatology such as hirsutism. Decrease in insulin itself should also result in a decrease or modification in risk factors associated with hyperinsulinism and insulin resistance, including the development of type 2 diabetes, dyslipidemias, hypertension, and cardiovascular disease.
Most recently, insulin resistance has been demonstrated in girls with premature adrenarche (PA) (18, 19). PA, defined by the onset of pubic hair in girls before age 8 yr, is associated with adrenal androgen levels higher than expected for age. It has been suggested that PA may be an antecedent of PCOS (18, 20, 21). Indeed, although PA was once considered a benign condition, it is now known to be associated with risk for many of the same conditions as PCOS, including hirsutism, obesity, AN, hyperinsulinism, and insulin resistance (22, 23, 24). Of relevance is the hypothesis proposed by Barker et al. (25) that lower birth weight or reduced fetal growth is related to the development of type 2 diabetes mellitus, hypertension, and hyperlipidemia, as well as "Syndrome X" in adult life (26). Ibanez et al. (27, 28), a major proponent of an extension of this theory, state in this issue of the journal that a "constellation of hirsutism, hyperandrogenism, oligomenorrhea, dyslipidemia and hyperinsulinism in lean, young women may already be a late stage of a developmental disorder starting early in life" (29). Germane to the current report is the prior work by Apter et al. (30), in which hyperinsulinemic insulin resistance and reduced IGFBP-I and SHBG levels were associated with ovarian hyperandrogenism in young adolescent girls, suggesting that peripubertally insulin plays a contributory role to the development of the metabolic derangement ascribed to PCOS.
In the current report, Ibanez et al. (29) describe 10 "nonobese adolescents ... in whom the appearance of ovarian hyperandrogenism was heralded by premature pubarche, hyperinsulinism and dyslipidemia before puberty and, even earlier, by a low birth weight," in whom treatment with metformin, an insulin sensitizing agent, resulted in a decrease in hirsutism score, insulin response to an oral glucose tolerance test (OGTT), free androgen index, and androgen response to stimulation with GnRH. Serum triglyceride, total cholesterol, and LDL cholesterol levels declined whereas HDL cholesterol increased. Menses occurred within 4 months of treatment. Perhaps the most important aspect of this study is that metformin use was addressed in a young and lean population at risk for the development of cardiovascular disease. If these findings are reproduced in larger placebo and weight-controlled studies, significant implications could be drawn including the encouragement of the use of an insulin-sensitizing agent, like metformin, in young patients with hyperinsulinemia, particularly those with a family history of type 2 diabetes and cardiovascular disease.
At the present time, few studies have addressed the use of metformin in children as a treatment for either insulin resistance or obesity. The action of metformin is not fully known. It inhibits hepatic glucose production and increases peripheral tissue sensitivity to insulin. It is thought to decrease insulin action on both the ovary and the liver. In vitro, therapeutic concentrations of metformin have been shown to simulate the tyrosine kinase activity of the intracellular portion of the ß-subunit of the human insulin receptor, but higher levels of metformin inhibited the tyrosine kinases (31). The most common morbity associated with its use is gastrointestinal distress, specifically diarrhea and abdominal pain, which is often transient and seems to be lessened if the dose is gradually increased. Although patients with renal insufficiency seem to be most at risk of developing severe lactic acidosis after receiving metformin, a recent case was observed in an elderly patient with normal renal function (32).
Table 1 cites a limited number of recent studies that have assessed the efficacy of metformin as an agent capable of reversing metabolic and ovarian abnormalities often associated with increased levels of insulin.
Although not reported by all investigators, metformin seems to cause a decline in insulin levels, an increase in insulin sensitivity, and a decrease in LH, androgens, and lipid levels and the ratio of IGF-I/IGFBP-I. Many of these changes occur even in the absence of changes in body mass index (BMI). It may not been effective in profoundly obese women. It has been associated with modest declines in weight and increases in normal menstrual cycling patterns and ovulation. It has been used during pregnancy with safety. As seen in Table 1, use of metformin in adolescents has been extremely limited although its use seems to be safe in small pilot studies.
Comparison of the current study by Ibanez et al. (29) with the reports above suggests that the dose (1275 mg/day) and length of time (6 months) of metformin used is similar to the previous studies. The young mean age of the girls was 16.8 yr (range, 13–20; most were 15–17 yr, with one only 20 yr of age; personal communication) makes this of particular interest because efforts aimed at preventing long-term cardiovascular complications of hyperinsulinism and modifying effects of androgen excess in at-risk adolescents and young women of child-bearing age may be speculated to be more effective if begun at a younger age. Noteworthy is the fact that these adolescents were all at least 3 yr postmenarche and had been documented to have hyperinsulinemia—8 of 10 prepubertally at diagnosis of premature pubarche—and in early puberty, with mean serum insulin levels 2 or more SD scores above the mean for age. Thus, identification of an at-risk population would provide a younger-aged population in whom potential benefit from targeted therapy with insulin-sensitizing agents like metformin could be achieved.
Unlike the majority of women in the prior studies where metformin has been effective, these adolescents were lean (mean BMI, 21.9 kg/m2) and not obese. Other efforts aimed at reducing insulin levels in the Ibanez et al. (29) cohort, such as weight reduction, were, thus, not a real option. The significant decrease in hirsutism is noteworthy because other agents such as birth control pills or antiandrogens, which have been used to reduce the effects of hyperandrogenism in women with hyperinsulism and PCOS, have often resulted in limited success (52, 53). Although the skin findings of AN are frequently observed in obese women with insulin resistance and obesity, only one of these adolescents had mild AN (personal communication), perhaps because of their lean BMI. However PAI-1, another marker known to be associated with increased cardiovascular vascular events and insulin-resistance in obese and type 2 diabetic patients, was elevated and subsequently decreased with metformin use (54). Furthermore, although her population was not small for gestational age (i.e. weight less <10%), the mean birth weights were in the lowest quartile for gestational age (55), corroborating the previous reports linking low birth weight to insulin resistance in adulthood (25, 27).
As in many small initial clinical drug studies, there was no placebo control group; however, the reversal of improvements on withdrawal of metformin was impressive. Indeed, although this report is of a very small group of patients and may, in fact, represent a subgroup of girls at risk because of low birth weight, it supports the hypothesis that hyperinsulinemia may have an early critical role in the development of PCOS and "Syndrome X" and challenges us to confirm and expand these findings. Clearly, targeted treatment of a younger population of girls at risk for cardiovascular disease and type 2 diabetes associated with hyperinsulinism such as precocious pubarche may attenuate, decrease, or even prevent progression to PCOS in the reproductive years.
The public health ramification of this report may be quite significant. Currently, the incidence of type 2 diabetes in adolescents is increasing at an alarming rate (56). In Tokyo, Japan, for example, type 2 diabetes accounts for close to 80% of all childhood diabetes [Dr. K. Kida (Ehime University, Matsuyama, Japan), personal communication]. Is there sufficient safety and efficacy data available for metformin to suggest it be used in young patients at-risk for the development of type 2 diabetes?
As with any new therapeutic model, long-term safety issues need to be addressed. In the current study, minimal transient gastrointestinal discomfort was noted in three of the adolescents with no report of hypoglycemia or lactic acidosis. Use of insulin sensitizers must be carefully monitored as most recently evidenced by the Food and Drug Administration withdrawal of the thiazolidinedione troglitazone due to fatal liver toxicity associated with its use. At the present time, given the lack of established long-term efficacy and safety in either adults, adolescents, and certainly children, until prospective large-scale clinical trials are performed, it would seem prudent to limit the use of metformin to adolescents at risk for the development of PCOS and type 2 diabetes to clinical trials. The report by Ibanez et al. (29) should encourage initiation of such studies.
|
Accepted August 13, 2000.
|
References |
|---|
|
TopIntroductionReferences |
|---|
as a cause of polycystic ovarian
syndrome. Fertil Steril. 53:785–791.[Medline]
activity and serum free testosterone
after reduction of insulin secretion in polycystic ovary syndrome. N Engl J Med. 335:617–623. activity and serum androgens. J
Clin Endocrinol Metab. 82:4075–4079.-hydroxycorticosteroid intermediates
response to adrenocorticotropin in hyperandrogenic women: apparent
relative impairment of 17,20-lyase activity. J Clin Endocrinol
Metab. 81:881–886.[Abstract]
response to human chorionic gonadotrophin in women with insulin
resistance-related polycystic ovary syndrome. Hum Reprod. 15:21–23.This article has been cited by other articles:
 |
|
 |
|
  J. Sheeder, S. H. Travers, and C. Stevens-Simon Is This Patient Insulin Resistant? How Much Does It Matter? Clinical Pediatrics, November 1, 2003; 42(9): 835 - 839. [PDF]
|
|
|
|
 |
|
 S. A. Arslanian, V. Lewy, K. Danadian, and R. Saad Metformin Therapy in Obese Adolescents with Polycystic Ovary Syndrome and Impaired Glucose Tolerance: Amelioration of Exaggerated Adrenal Response to Adrenocorticotropin with Reduction of Insulinemia/Insulin Resistance J. Clin. Endocrinol. Metab., April 1, 2002; 87(4): 1555 - 1559. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Palmert, C. M. Gordon, A. I. Kartashov, R. S. Legro, S. J. Emans, and A. Dunaif Screening for Abnormal Glucose Tolerance in Adolescents with Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., March 1, 2002; 87(3): 1017 - 1023. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Ibanez, K. Ong, N. Potau, M. V. Marcos, F. de Zegher, and D. Dunger Insulin Gene Variable Number of Tandem Repeat Genotype and the Low Birth Weight, Precocious Pubarche, and Hyperinsulinism Sequence J. Clin. Endocrinol. Metab., December 1, 2001; 86(12): 5788 - 5793. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Ibanez, C. Valls, A. Ferrer, M. V. Marcos, F. Rodriguez-Hierro, and F. de Zegher Sensitization to Insulin Induces Ovulation in Nonobese Adolescents with Anovulatory Hyperandrogenism J. Clin. Endocrinol. Metab., August 1, 2001; 86(8): 3595 - 3598. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
