Administration of Metformin to a Diabetic Woman with Extreme Hyperandrogenemia of Nontumoral Origin: Management of Infertility and Prevention of Inadvertent Masculinization of a Female Fetus

Nicholas J. Sarlis, Stacie J. Weil and Lawrence M. Nelson

Developmental Endocrinology Branch (S.J.W., L.M.N.), National Institute of Child Health and Human Development; and Molecular and Cellular Endocrinology Branch (N.J.S.), National Institute of Diabetes, Digestive, and Kidney Diseases, National Institutes of Health, Bethesda, Maryland 20892

Address all correspondence and requests for reprints to: Dr. Nicholas J. Sarlis, Molecular and Cellular Endocrinology Branch/National Institute of Diabetes, Digestive, and Kidney Diseases, National Institutes of Health, Building 10, Room 8D12C, 10 Center Drive-MSC 1758, Bethesda, Maryland 20892-1758. E-mail: njsarlis{at}box-n.nih.gov


    Introduction
 Top
 Introduction
 Clinical Presentation
 Discussion
 References
 
The polycystic ovary syndrome (PCOS) is characterized by chronic hyperandrogenemic anovulation and is considered to be a common cause of infertility because it affects 5–15% of premenopausal women (1, 2, 3). Metformin has been shown, in several studies, to significantly lower androgen levels and initiate menstrual cyclicity in women with PCOS (4, 5, 6, 7, 8). The mechanism of action of metformin in PCOS is unknown, but it is believed to involve the combined effects of decreased insulin action on the ovary (lowering ovarian androgen production rates) and on the liver [resulting in an increase in sex hormone binding globulin (SHBG) levels] (4, 9, 10, 11).

Though metformin is effective in women with PCOS with mild to moderate elevations of androgen serum levels (2, 4, 5, 6, 7, 8), its effects in the considerably smaller subset of PCOS cases associated with extreme hyperandrogenemia [testosterone (T) > 200 ng/dL (6.9 nmol/L)] and maternal virilization have not been studied. Such extreme levels of hyperandrogenemia during pregnancy are usually caused by an ovarian tumor or luteoma of pregnancy and can be treated effectively by surgery (12). However, this is not the case when these high levels are caused by PCOS. In fact, although the possibility of fetal masculinization exists in the presence of severe maternal hyperandrogenism, there are presently no effective preventive measures when this is found in women with PCOS (13).


    Clinical Presentation
 Top
 Introduction
 Clinical Presentation
 Discussion
 References
 
In August 1995, a 29-yr-old woman, gravida 2 para 0, presented for evaluation of virilization and infertility. After an unremarkable adolescence and history of two elective early abortions at age 19 and 20, the patient, at age 22, had noted the onset of rapidly worsening hirsutism, cephalic male-pattern baldness, deepening of her voice, and an increase in muscle mass. By age 25, she was shaving her facial hair twice daily and had experienced a 30-kg increase in weight over the proceeding 3 yr. The patient had presented with oligomenorrhea at age 23, and had been amenorrheic since age 24. Thereafter, menstruation occurred only after progesterone-induced withdrawal. The patient also had a decade-long history of mild hypertension. On presentation, she was an obese black woman with an android body habitus. Her height was 168.8 cm; weight, 88.5 kg; body mass index, 31.1 kg/m2; and blood pressure, 148/93 mm Hg. The patient was hirsute and had clitoromegaly and acanthosis nigricans. She had markedly elevated serum androgen levels (Fig. 1Go) and a normal fasting blood glucose. Baseline serum TSH and PRL, an ACTH stimulation test for serum cortisol, and 24-h excretion rates for urinary free cortisol and 17-hydroxysteroids were within normal limits. Computed tomography and ultrasound evaluation of the abdomen and pelvis showed no evidence of tumor. The patient was treated with oral contraceptives and was reevaluated in February 1996. By that time, she had demonstrated a significant reduction in serum androgen levels (Fig. 1Go). The diagnosis of PCOS was established.



View larger version (24K):
[in this window]
[in a new window]
 
Figure 1. Free and total T levels over time. The range of normal values (NIH laboratory) for free T in both nonpregnant and pregnant women are 3–19 pg/mL (10.5–66.0 pmol/L). The excursions in total T levels exactly paralleled these seen in free T levels, with a maximum pre-pregnancy value of 432 ng/dL (15.0 nmol/L) in August 1995 and a dramatic response to metformin therapy. The range of normal values for total T for nonpregnant women is 20–80 ng/dL (0.70–2.78 nmol/L). MET, metformin; OCP, oral contraceptive pill.

 
Subsequently, oral contraceptives were discontinued because the patient desired a pregnancy. She was placed on a weight reduction regimen, including a short period of treatment with phentermine/fenfluramine combination, but was unable to lose weight. Three months after the discontinuation of oral contraceptives, she remained amenorrheic. In October 1996, the patient again had significantly elevated androgen levels (Fig. 1Go) and was now meeting diagnostic criteria for diabetes mellitus, by an oral glucose tolerance test (despite normal fasting blood glucose levels). She began metformin (500 mg p.o. b.i.d.) and continued this medication for 5 months, with resultant marked decreases of serum androgens. Within 2 months after starting metformin, the patient was experiencing normal menstrual cycles.

In February 1997, the patient discontinued the metformin therapy on her own, and she became pregnant within a month. In June 1997, at 10.5 weeks of gestation, a relapse of marked hyperandrogenemia was documented (Fig. 1Go). Because of the possibility of masculinization of a female fetus by this degree of excess androgens, and after extensive counseling with regard to the possible risks and benefits of metformin administration during pregnancy, the patient was again placed on metformin, 500 mg p.o. b.i.d. at 14 weeks of gestation (August 1997). Both total and free T levels decreased dramatically (Fig. 1Go). At 29.5 weeks of gestation, the patient developed preeclampsia. Because of a rapid deterioration of her condition, primary cesarean section was performed at 30 weeks of gestation. A nonmasculinized 1,335-g female infant was delivered. The patient discontinued the metformin therapy in the immediate postpartum period for 3 months (while she was nursing her infant). During that time, she maintained normal blood glucose levels. After the discontinuation of breast-feeding, the patient was again placed on metformin, 500 mg p.o. b.i.d.; and she continues to take this medication to date, with excellent metabolic response (normalization of total and free T levels and glycemia, and resumption of menses).


    Discussion
 Top
 Introduction
 Clinical Presentation
 Discussion
 References
 
Metformin has been shown, in several studies, to significantly lower androgen levels in women with PCOS and to initiate menstrual cyclicity (4, 5, 6, 7, 8). These findings were not reproduced in one recent study of morbidly obese women with PCOS (14). In this study, neither plasma androgen or insulin levels decreased significantly after metformin therapy, probably because of the overwhelming metabolic effect(s) of extreme obesity in the study population (average body mass index, 40.0 kg/m2). Another study was also unable to show a metformin-induced decrease in indices of insulin resistance in patients with PCOS (15). However, resumption of menses shortly after initiation of metformin therapy has been reported in a significant portion of patients with PCOS (4, 5, 6, 7, 8). It is believed that metformin leads to decreases in serum free androgen levels, thus permitting resumption of menstrual cyclicity (2, 4).

Our case raises two important issues relevant to the presence of extremely high androgen levels (associated with maternal virilization) during pregnancy: 1) the indeterminate, yet existent, risk of masculinization of a female fetus; and 2) the risk of metformin therapy during pregnancy. With regard to issue 1, the possibility of fetal masculinization exists in maternal congenital adrenal hyperplasia, maternal ovarian tumors, exogenous androgen administration to the mother, luteoma of pregnancy, and (rarely) in placental aromatase deficiency and in mothers with PCOS (12, 13, 16, 17, 18, 19). The exact incidence of this side effect of maternal androgens varies greatly among case reports and published small series (12, 20, 21). Excess androgen exposure during fetal development is relevant not only to genital differentiation but also to central nervous system functional mechanisms underlying subsequent expression of sociosexual behavior and gender identity (22, 23). Whereas fetal congenital adrenal hyperplasia is the most common cause of ambiguous genitalia in female neonates, luteoma of pregnancy is the most frequent cause for maternal virilization during pregnancy, as approximately 50–60% of female fetuses born to virilized mothers exhibit signs of masculinization (20). Interestingly, virilization of the fetus has not proven to be tightly correlated with maternal androgen levels. In general, such a risk is increased if the mother herself shows clinically significant manifestations of virilization (12). The timing of androgen exposure during gestation has been shown to affect the neonatal phenotype, as expected (12, 13).

With regard to issue 2 (i.e. the effects of metformin during pregnancy), metformin has been rated by the Food and Drug Administration as a category B medication (24). Mouse embryos exposed to metformin at doses of 500–2,550 mg daily have shown no major malformations in the offspring (25). A South African group has published several series describing women with type II diabetes mellitus on metformin therapy throughout pregnancy; no adverse effects were reported, with regard to fetal and neonatal development (26).

In conclusion, we report a case where maternal hyperandrogenism during pregnancy was not attributable to common causes, i.e. ovarian tumor or luteoma of pregnancy (which can all be treated effectively, either medically or surgically), but rather was due to PCOS. This case is of additional interest because of significant increases in maternal androgen levels at the early stages of pregnancy, a phenomenon that is quite rare in normal pregnancy (27, 28) but may be more frequent in PCOS (13); there is a paucity of information with regard to this phenomenon in women with PCOS. Our case suggests that metformin may be effective for anovulatory women with PCOS and extreme hyperandrogenemia. These patients also might be potentially maintained on metformin throughout pregnancy, aiming at the prevention of inadvertent fetal masculinization (feto-protective treatment).

Received December 9, 1998.

Revised January 12, 1999.

Accepted February 3, 1999.


    References
 Top
 Introduction
 Clinical Presentation
 Discussion
 References
 

  1. Hull MG. 1987 Epidemiology of infertility and polycystic ovarian disease: endocrinological and demographic studies. Gynecol Endocrinol. 1:235–240.[Medline]
  2. Nestler JE. 1997 Role of hyperinsulinemia in the pathogenesis of the polycystic ovary syndrome and its clinical implication. Semin Reprod Endocrinol. 15:111–117.[Medline]
  3. Speroff L. 1994 Anovulation and the polycystic ovary. In: Speroff L, Glass R, Kase N, eds. Clinical gynecology endocrinology and infertility. Baltimore: William & Wilkins; 457–482.
  4. Nestler JE, Jakubowicz DJ, Evans WS, et al. 1998 Effects of metformin on spontaneous and clomiphene-induced ovulation in the polycystic ovary syndrome. N Engl J Med. 338:1876–1881.[Abstract/Free Full Text]
  5. Velazquez E, Mendoza S, Hamer T, Sosa F, Glueck C. 1994 Metformin therapy in polycystic ovary syndrome reduces hyperinsulinemia, insulin resistance, hyperandrogenism, and systolic blood pressure, while facilitating normal menses and pregnancy. Metabolism. 43:647–654.[Medline]
  6. Velazquez E, Acosta A, Mendoza S. 1997 Menstrual cyclicity after metformin therapy in polycystic ovary syndrome. Obstet Gynecol. 90:392–395.[Abstract/Free Full Text]
  7. Diamanti-Kandarakis E, Kouli C, Tsianateli T, Bergiele A. 1998 Therapeutic effects of metformin on insulin resistance and hyperandrogenism in polycystic ovary syndrome. Eur J Endocrinol. 138:269–274.[Medline]
  8. Morin-Papunen LC, Koivunen RM, Ruokonen A, Martikainen HK. 1998 Metformin therapy improves the menstrual pattern with minimal endocrine and metabolic effects in women with polycystic ovary syndrome. Fertil Steril. 69:691–696.[CrossRef][Medline]
  9. Nestler J, Jakubowicz D. 1996 Decreases in ovarian cytochrome P450c17 activity and serum free testosterone after reduction of insulin secretion in polycystic ovary syndrome. N Engl J Med. 335:617–623.[Abstract/Free Full Text]
  10. Nestler JE, Jakubowicz DJ. 1997 Lean women with polycystic ovary syndrome respond to insulin reduction with decreases in ovarian P450c17-alpha activity and serum androgens. J Clin Endocrinol Metab. 82:4075–4079.[Abstract/Free Full Text]
  11. Laatikainen TJ, Apter DL, Paavonen JA, Wahlstrom TR. 1980 Steroids in ovarian and peripheral venous blood in polycystic ovarian disease. Clin Endocrinol (Oxf). 13:125–129.[Medline]
  12. McClamrock HD, Adashi EY. 1992 Gestational hyperandrogenism. Fertil Steril. 57:257–270.[Medline]
  13. Ben-Chetrit A, Greenblatt EM. 1995 Recurrent maternal virilization during pregnancy associated with polycystic ovarian syndrome: a case report and review of the literature. Hum Reprod. 10:3057–3062.[Abstract]
  14. Ehrmann D, Cavaghan M, Imperial J, Sturis J, Rosenfield R, Polonsky K. 1996 Effects of metformin on insulin secretion, insulin action and ovarian steroidogenesis in women with polycystic ovary syndrome. J Clin Endocrinol Metab. 82:524–530.[Abstract/Free Full Text]
  15. Acbay O, Gundogdu S. 1996 Can metformin reduce insulin resistance in polycystic ovary syndrome? Fertil Steril. 65:946–949.[Medline]
  16. Shozu M, Akasofu K, Harada T, Kubota Y. 1991 A new cause of female pseudohermaphroditism: placental aromatase deficiency. J Clin Endocrinol Metab. 72:560–565.[Abstract]
  17. Magendantz HG, Jones DED, Schomberg DW. 1972 Virilization during pregnancy associated with polycystic ovary disease. Obstet Gynecol. 40:156–162.[Medline]
  18. Fayez JA, Bunch TR, Miller GL. 1974 Virilization in pregnancy associated with polycystic ovary disease. Obstet Gynecol. 44:511–521.[Medline]
  19. Bilowus M, Abassi V, Gibbons MD. 1986 Female pseudohermaphroditism in a neonate born to a mother with polycystic ovarian disease. J Urol. 136:1098–1100.[Medline]
  20. Manganello P, Adams L, Harris R, Ornvold K. 1995 Virilization during pregnancy with spontaneous resolution postpartum: a case report and review of the English literature. Obstet Gynecol Surv. 50:404–410.[CrossRef][Medline]
  21. Shortie BE, Warren MP, Tsin D. 1987 Recurrent androgenicity in pregnancy. A case report and literature review. Obstet Gynecol. 70:462–464.[Abstract]
  22. Michael R, Zumpe D, Bonsail R. 1992 The interaction of testosterone with the brain of the orchidectomized primate fetus. Brain Res. 570:68–74.[CrossRef][Medline]
  23. Sato T, Koizumi S. 1991 Effects of fetal androgens on childhood behavior. Acta Paediatr Jpn. 33:639–644.[Medline]
  24. Anonymous. 1998 Metformin. In: Physicians desk reference. Montvale, NJ: Medical Economics Co., Inc.; 795–800.
  25. Denno K, Sadler T. 1994 Effect of biguanide class of oral hypoglycemic agents on mouse embryogenesis. Teratology. 49:260–266.[Medline]
  26. Coetzee E, Jackson W. 1980 Pregnancy in established noninsulin-dependent diabetics. A five-and-a-half year study at Groote Schuur Hospital. S Afr Med J. 58:795–802.[Medline]
  27. Mizuno M, Lobotsky J, Lloyd CW, Kobayashi T, Murasawa Y. 1968 Plasma androstenedione and testosterone during pregnancy and in the newborn. J Clin Endocrinol Metab. 28:1133–1142.[Medline]
  28. Bammann BL, Coulan CB, Jiang N-S. 1980 Total and free testosterone during pregnancy. Am J Obstet Gynecol. 137:293–298.[Medline]