CHICAGO – When women with polycystic ovary syndrome (PCOS) took metformin during pregnancy, preterm births and late miscarriages were reduced, but researchers saw no effect on blood glucose levels or other diabetes-related maternal outcomes.

A Nordic study of pregnant women, dubbed PregMet 2, followed the course of 487 women with PCOS who were randomized to take 2,000 mg of metformin or placebo daily during pregnancy.

The question that PregMet 2 sought to answer, said Tone Løvvik, MD, speaking at the annual meeting of the Endocrine Society, was “Can metformin prevent late miscarriages and preterm birth in women with PCOS?” The study’s results answered that question in the affirmative, with some caveats.

Among the entire study population, those with PCOS who took placebo had a 9.6% incidence of late miscarriage or preterm birth. For those taking metformin, the figure was 5%, similar to the 5.2% risk seen in the population-wide Norwegian Birth Registry, said Dr. Løvvik, of the Norwegian University of Science and Technology, Trondheim.

This analysis yielded a number needed to treat (NNT) for benefit of metformin of 22, she said.

However, Dr. Løvvik said that “the elephant in the room” is the lack of benefit of metformin for rates of gestational diabetes, hypertension in pregnancy, or preeclampsia in PregMet 2.

Of the various endpoint analyses, “the most surprising and striking one is the absolute lack of effect on gestational diabetes in this high-risk population,” said Dr. Løvvik. “There’s not even a tendency towards effect. … Metformin had no effect on prevention, treatment, or the need for additional insulin, indicating it was actually as effective as placebo.”

Considering that “metformin is now a part of the standard treatment for gestational diabetes, according to many national guidelines, we think that it’s questionable that it’s never, ever been tested against placebo for this diagnosis,” said Dr. Løvvik.

Pregnant women with PCOS have an increased risk of complications, and metformin is often prescribed off label to attempt to address some of these complications. Some previous work had shown metformin to be helpful, but previous studies have been underpowered, said Dr. Løvvik.

PregMet 2’s protocol attempted to address the literature gap; over 3 years, it called for enrollment of 1,000 women in the first trimester of pregnancy, randomized 1:1 to receive metformin or placebo in pregnancy.

At that enrollment level, the study would aim to show a 50% reduction in late miscarriages and preterm birth, with a power of 85%. Two predetermined analysis were planned at the end of the 3-year study, one for the intent-to-treat population and one for the per-protocol group.

In the end, however, just 487 patients were enrolled and randomized over a period of 5 years, despite recruitment at 14 centers in Iceland, Norway, and Sweden.

Richard Legro, MD, an ob.gyn. from Pennsylvania State University, Hershey, was in attendance. While praising the overall high quality of the study, he commented, “The primary concern is that you stopped the study when you hadn’t even achieved half your sample size.”

Dr. Løvvik said that a variety of factors contributed to the low study enrollment. One primary issue was that participation required additional visits on the part of participants; the study was not part of routine prenatal care, she said. Also, the placebo used in the study expired before enrollment was completed, complicating execution of the study.

She also noted that nearly a quarter of patients had a history of depression, and one in five had a history of migraine. More than half of patients overall had at least one previous or ongoing chronic medical condition.

A small number of patients in each group were excluded for protocol violations, pregnancy termination, intrauterine fetal demise, or loss to follow-up. The final intent-to-treat population included 238 in the metformin group and 240 who received placebo.

The per-protocol analysis was conducted two ways: In the first analysis, only those who dropped out were excluded, leaving 209 in the metformin group and 223 in the placebo group. The second analysis included those whose adherence to taking the study medication was assessed at 90% or better. For this second per-protocol analysis, just 142 metformin takers and 156 in the placebo group remained.

In the intent-to-treat population, the primary endpoint – a composite outcome of late miscarriage and preterm delivery – yielded an odds ratio of 1.99 favoring metformin over placebo, a figure that fell short of statistical significance (95% confidence interval, 0.93-4.51; P = .08).

However, there was significantly less weight gain in those taking metformin in this population (8.7 vs. 11.48 kg; P less than .0001). And newborns had significantly larger head circumferences if their mothers with PCOS took metformin (35.4 vs. 35.0 cm; P = .006).

Dr. Løvvik and her colleagues subsequently performed the same analysis on the first per-protocol group, one that excluded the 46 patients who were early dropouts after randomization. Here, the odds ratio favoring metformin for the composite primary outcome measure was a significant 2.55 (n = 432; 95% CI, 1.10-6.42; P = .03). Results for the final per-protocol analysis that excluded dropouts and those with less than 90% adherence to study medication were similar, with an odds ratio of 2.76 in favor of metformin use during pregnancy (n = 298; 95% CI, 1.0-8.82; P = .05).

In response to an audience question, Dr. Løvvik said that, though a per-protocol analysis of some sort had been predetermined, the decision to perform the narrower analysis on the highly adherent group was made later. However, compared to thresholds for adherence in other studies, “ours was way higher – so it’s too strict,” she acknowledged.

At baseline, the median age overall was about 30 years, and the median body mass index was about 28 kg/m². Patients were at a median 74 days gestational age of pregnancy at randomization, and 53%-61% of patients were on metformin at the time of conception, with no significant difference between the metformin and placebo group.

A little less than half (40%-46%) of participants had a spontaneous conception, and most participants (56%-59%) had no prior term deliveries. Participants were overwhelmingly (91%-97%) white and Nordic.

There were no serious maternal, fetal, or neonatal safety signals in either study group.

The study followed the earlier PregMet study, which also tracked use of metformin in pregnancy for women with PCOS.

koakes@mdedge.com

Metformin reduces preterm births, late miscarriages in PCOS