The Effects of Metformin on Pregnancy and Miscarriage Rates in Polycystic Ovary Syndrome (PCOS)

Women with PCOS represent about 5-10% of the general female population and one third of the women treated for infertility. Thus, the development of new therapies to improve the efficiency of ovulation induction treatments and the outcome of pregnancy, and to reduce the long-term risks of the syndrome would bring important health benefits.

The central role played by insulin resistance and hyperinsulinemia in PCOS - causing hyperandrogenism, premature follicular atresia, anovulation, oligo-amenorrhea and anovulatory infertility - has led to the use of insulin-lowering drugs for the treatment of this syndrome. The most studied agent is metformin, a biguanide antihyperglycemic drug used to treat Type 2 diabetes mellitus. It has been shown to improve significantly hyperinsulinemia and insulin resistance, to decrease androgen levels, and to improve menstrual pattern and, alone or in addition to clomiphene citrate, to induce ovulation and improve pregnancy rates in women with PCOS in some studies (1,2). Metformin may also decrease risks of early spontaneous miscarriage and gestational diabetes in PCOS (3-6). Two recent RCTs, however, have shown no beneficial effect of metformin compared to placebo as regards rates of pregnancy, miscarriage or life births in women with PCOS (7,8).

Our hypothesis is that metformin may improve pregnancy rates and decrease miscarriage occurrence and complications of pregnancy, such as toxemia and gestational diabetes, in women with PCOS. This multicenter randomized placebo-controlled study is conducted in all five University Hospitals of Finland (Oulu, Kuopio, Helsinki, Tampere and Turku). Blood samples are drawn and the oral glucose tolerance test (OGTT) done before and at 3 months of treatment, after which the treatment with placebo/metformin is continued another 6 months' period together with the appropriate infertility treatment. If pregnancy occurs, the OGTT is done at 7-8 weeks of pregnancy and the placebo/metformin treatment is continued until 12 weeks of pregnancy. The study has already started and is estimated to continue at least until the end of 2009. Power analysis indicated that a minimum of 60 pregnant patients are needed in each group to decrease the risk of miscarriage from 44% to the normal 15%.