Metformin use in pregnancy: promises and uncertainties
Metformin use in pregnancy: promises and uncertainties
Robert S. Lindsay 0 1 2
Mary R. Loeken 0 1 2
0 Section on Islet Cell and Regenerative Biology, Joslin Diabetes Center , Boston, MA , USA
1 Institute of Cardiovascular and Medical Sciences, British Heart Foundation (BHF) Glasgow Cardiovascular Research Centre, University of Glasgow , 126 University Place, Glasgow G12 8TA , UK
2 Robert S. Lindsay
Metformin has been prescribed in pregnancy for over 40 years; for much of this time, use has been limited both in numbers and geographically, and the evidence base has been confined to observational studies. In early years, perceived safety concerns and lack of availability of the drug in many countries acted as a barrier to use. More recently, RCTs have begun to examine the role of metformin in pregnancy in much-needed detail. However, this evidence base has been interpreted differently in different countries, leading to very wide variation in its current application in pregnancy. In this short review, we will discuss the history of metformin in pregnancy and highlight some of the key clinical trials. We will then consider some of the remaining controversies associated with metformin use in pregnancy, most important of these being the potential for long-term 'programming' effects on the fetus as a result of metformin being able to cross the placenta. We will also consider clinical situations where metformin might be avoided. Finally, we will discuss some future directions for this drug as it reaches its sixtieth anniversary.
Metformin; Pregnancy; Review; Teratogenesis
-
The initial development and use of metformin (outside of
pregnancy) are reviewed elsewhere in this issue of
Diabetologia [
1
]. With regard to pregnancy, it is important
to note that it was acknowledged very early on that metformin
crossed the placenta. More recent studies show similar plasma
concentrations in the maternal and fetal circulation [
2
].
Further, the combination of increased lactic acidosis risk
(mainly observed with the metformin-related biguanide,
phenformin) and the relatively hypoxic fetal environment led
to important concerns regarding potential adverse effects of
metformin use in pregnancy, for both mother and child. In
fact, the safety concerns related to phenformin use resulted
in the withdrawal of metformin in many, although not all,
countries [
3
].
These early concerns are charted in influential reports of
the Aberdeen International Colloquia on Carbohydrate
Metabolism in Pregnancy and the Newborn. The first
colloquium, reported in 1975, included an entire chapter on ‘the use
of sulphonylureas, biguanides and insulin in pregnancy’ [
4
].
By the time of the fourth report in 1988, the topic of use of
metformin was given only a few lines and it was noted that use
was not widespread [
5
].
Metformin use did, however, continue in other parts
of the world. In developing countries, the relatively low
cost of metformin compared with insulin made it an
attractive option. Coetzee and colleagues published a
series of important observational papers, commencing
in the late 1970s, examining the use of metformin in
South Africa [
6–8
]. In South Africa and other countries,
where metformin was routinely used to treat type 2
diabetes, exposure inevitably began to occur in early
pregnancy leading to the separate analysis of safety in
early pregnancy, particularly regarding miscarriage and
congenital anomaly.
However, it was as metformin use became more popular in
polycystic ovarian syndrome (PCOS) that a more robust
literature developed, investigating exposure of the fetus to
metformin in early pregnancy and, therefore, safety of its use.
Safety and efficacy of metformin during pregnancy
Safety in early pregnancy The early literature regarding
metformin use in early pregnancy in humans was based
on observational findings and of variable quality.
Studies were usually small and it was often difficult to
tease out the potential teratogenic effects of metformin,
particularly as opposed to the well-established effects of
maternal hyperglycaemia to increase risk of congenital
malformation [
9
]. Similarly, animal studies have not
been completely conclusive, and while increases in
embryonic AMP-activated protein kinase (AMPK; AMPK
activation being one potential effect of metformin) may
be key in diabetic embryopathy, animal studies have not
suggested an increase in embryopathy with early
metformin exposure in vivo [
10
].
More recently, a meta-analysis based on metformin
exposure in 351 women with PCOS has been carried
out. Interestingly, the findings of this study do not
suggest an increase in congenital anomaly with metformin
use in pregnancy (OR of major birth defect 0.86 [95%
CI 0.18, 4.08]) [
11
]. However, this estimate is based on
a small sample size and, therefore, the confidence
intervals remain wide.
Metformin has also been extensively analysed in the
context of PCOS-associated miscarriage and pregnancy
indu (...truncated)