Two new pieces of research have provided further evidence supporting the use of metformin in pregnant women with diabetes, even after no benefit was found for the primary outcome of interest.
Metformin might not reduce the likelihood of developing gestational diabetes mellitus, but there are still other benefits on offer.
Although metformin is used increasingly in the treatment of gestational diabetes mellitus and other high metabolic risk pregnancies (including in individuals with polyendocrine metabolic ovarian syndrome), there is yet to be a singular trial with enough statistical power to assess GDM as a primary endpoint.
To overcome this research gap, an international consortium has undertaken a patient-level meta-analysis from randomised trials that assessed the effect of metformin administered during pregnancy on maternal, birth, or neonatal outcomes. Their findings, published in NEJM Evidence, identified some unexpected associations.
Researchers collated individual-level data from 2297 participants across seven RCTs, 1159 of whom received metformin (the remaining 1138 received placebo). Of these, 18.3% were aged ≥35 years, 60.5% had PMOS, and 66.6% were obese (defined as a BMI ≥30). The mean gestational age at the start of treatment was 12.8 weeks, with daily metformin doses ranging between 500mg and 3000mg.
Adverse events affected a similar proportion of participants across the metformin and placebo groups: 73.8% of participants receiving metformin reported any side effect compared to 67.1% of participants who received placebo, with 7.5% and 5.1% of participants in each group permanently stopping treatment. Gastrointestinal side effects were most common, occurring in 66.3% and 57.0% of participants in the metformin and placebo groups respectively.
After accounting for the effects of maternal age, baseline BMI, and gestational age at trial enrolment, metformin was associated with reduced odds of developing gestational diabetes (based on the 2010 International Association of the Diabetes and Pregnancy Study Groups definition; odds ratio and 95% confidence interval 0.71 [0.52-0.98]), and experiencing low birth weight (0.66, 0.46-0.96) and gestational weight gain (-1.38, -1.86 to -0.90) compared to placebo.
“Metformin was not associated with reduced GDM using WHO 1999 or National Institute for Health and Care Excellence 2015 criteria,” the researchers wrote.
“Although IADPSG criteria are increasingly used in clinical practice, they include lower fasting blood glucose thresholds than earlier criteria, prompting ongoing debate regarding potential GDM overdiagnosis. The clinical importance of identifying and treating GDM at these lower thresholds remains controversial.”
There were also associations between metformin and maternal mean fasting glucose levels (mean difference and 95% CI -0.05, -0.09 to -0.01), neonatal head circumference percentile (2.43, 0.13-4.72), and gestational age at delivery in weeks (0.30, 0.06-0.54) compared to placebo.
There was no association observed for other maternal or neonatal outcomes following metformin treatment, including maternal fasting glucose, gestational hypertension, pre-eclampsia, the need for labour to be induced, neonatal hypoglycaemia, macrosomia, admission to the neonatal intensive care unit, respiratory distress, jaundice, and birth weight or length.
Related
The researchers concluded that metformin had limited clinical value in preventing GDM and associated adverse outcomes but that the role in reducing preterm birth warranted further exploration.
“The [finding that the] conventional benefits of metformin on diabetes prevention seen outside pregnancy were not replicated here [suggests] differential mechanisms in pregnancy that warrant further investigation,” they noted.
However, they were unclear on the mechanisms by which metformin could extend gestation.
“Hypotheses include improved placental function, utero-placental blood flow, and immunoregulation in pregnancy, especially in the insulin-resistant milieu of obesity and PMOS,” they wrote.
“The potential benefit of fewer preterm births with metformin warrants further study, including optimal timing for initiation and cost implications, as preterm delivery carries considerable neonatal risks, and the use of metformin in this context may prove worthwhile.”
Elsewhere, a small Australian retrospective observational cohort study has highlighted the potential benefits in using metformin in pregnant women with type 1 diabetes. The case series was published in the Internal Medicine Journal.
Eight pregnant women with type 1 diabetes (mean age 29.8 years, mean diabetes duration 17.0 years) were identified from three health services throughout Victoria. Five of the eight women were receiving multiple daily injections of insulin at the time of conception, with the remaining three using a continuous subcutaneous insulin infusion. The average preconception HbA1c level was 7.4%.
Metformin therapy began at an average gestational age of 25.5 weeks. Only one woman experienced nausea; the remainder did not report any adverse effects. The mean gestational age at delivery was 36.1 weeks, with an average birth weight of 3604g.
None of the neonates were deemed small for gestational age, although four were admitted to the NICU, three experienced respiratory distress syndrome, and two had neonatal hypoglycaemia. No major congenital abnormalities occurred.
Six of the eight women had continuous glucose monitoring data available, which allowed researchers to observe “notable improvements in glycaemic control” after metformin therapy was initiated.
The median time in range increased from 62.5% (interquartile range 59.5-65.5%) at baseline to 77% (71.3-82.8%) at time of delivery – a median absolute improvement of 13% and a median relative improvement of 117.8%. There was a corresponding decrease in the time above range, from 36.0% (31.3-37.8%) at baseline to 21.5% (17.5-29.3%) at the time of delivery.
“While individual responses varied and the cohort was relatively small, the overall pattern supports evidence for the role of metformin as an adjunctive therapy,” the researchers concluded. “In addition to improvements in CGM metrics, our data suggest that metformin may attenuate the progressive escalation in insulin requirements characteristic of advancing gestation.”
The researchers noted the need for larger, more controlled studies with longer-term neonatal follow-up to confirm the efficacy of metformin in pregnant women with type 1 diabetes as well as optimising the doses used in said patients.
