Oral Alternatives to Insulin for Gestational Diabetes?

The use of oral medications for glycemic control in pregnancy is controversial among experts. Nonetheless, glyburide and metformin are commonly prescribed for managing gestational diabetes mellitus (GDM) given their convenience and affordability. In early 2018 the American College of Obstetricians and Gynecologists (ACOG; Obstet Gynecol 2018; 131:e49) and the Society for Maternal-Fetal Medicine (SMFM; Am J Obstet Gynecol 2018; 218:B2) released guidelines on management of GDM. The two organizations agree on insulin as a preferred treatment for GDM when nutritional therapy has failed, but differ on the acceptability of oral agents as alternatives. This discrepancy has led to some confusion, which we will address in this discussion.

Use of glyburide for GDM has risen markedly during the past several years, but neither guideline recommends it as first-line therapy (with ACOG more strongly advising against its use). In a seminal trial of glyburide versus insulin for GDM that was published 18 years ago, glyburide was not detected in umbilical cord blood, and neonatal outcomes did not differ significantly between groups, although the study was not powered for infrequent outcomes (N Engl J Med 2000; 343:1134). Subsequent studies using more-sensitive glyburide assays suggested some in-utero exposure (Obstet Gynecol 2015; 125:583), which could be expected to promote endogenous insulin secretion, thereby leading to fetal overgrowth and neonatal hypoglycemia. Observational studies and meta-analyses also have raised concerns about excess risk for adverse neonatal outcomes with glyburide (JAMA Pediatr 2015; 169:452 and BMJ 2015; 350:h102).

In light of these uncertainties, a randomized trial was designed in France, and the results were recently published in JAMA. The study involved about 900 women with GDM diagnosed between 24 and 34 weeks' gestation and tested whether glyburide is noninferior to insulin regarding a neonatal composite outcome that included macrosomia, neonatal hypoglycemia, and hyperbilirubinemia (JAMA 2018; 319:1773). The incidence of this outcome was 27.6% (glyburide) versus 23.4% (insulin), a difference that did not meet criteria for noninferiority; in other words, this result leaves open the possibility that glyburide is inferior to insulin. It's worth noting that the failure of glyburide to achieve noninferiority was probably not related to maternal hyperglycemia, as glycemic control was on average better in the glyburide group than the insulin group; in fact, both maternal and neonatal hypoglycemia occurred significantly more frequently with glyburide than with insulin. These findings, published after both 2018 guidelines appeared, bolster ACOG's caution against glyburide.

What about metformin? Studies comparing risk for adverse outcomes with metformin versus insulin suggest at least equivalent performance of metformin regarding perinatal outcomes (e.g., Diab Med 2017; 34:27). Mothers receiving metformin seem to have less gestational weight gain and lower risk for pregnancy-induced hypertension, and their neonates have lower risk for severe hypoglycemia. However, treatment failure due to suboptimal control of hyperglycemia is relatively common, with about one-third of women requiring substitution of insulin for metformin to achieve glycemic targets.

Although first-trimester exposure to metformin is not expected to raise risk for congenital abnormalities (BMJ 2018; 361:k2477), long-term outcomes remain uncertain in offspring of mothers with GDM who received this agent, which is known to cross the placenta. Researchers are following a subgroup of children whose mothers were randomized to metformin or insulin during pregnancy as part of the Metformin in Gestational Diabetes trial (BMJ Open Diabetes Res Care 2018; 6:e000456). Results at one center indicate that at age 9 years, children born to mothers who received metformin are larger than those whose mothers received insulin (however, women in the metformin group show a trend toward higher body-mass index, so it's unclear to what extent the observed differences are due to this agent as opposed to factors associated with health behaviors). Unlike ACOG, SMFM considers metformin to be a first-line alternative to insulin for GDM, given the evidence for similar perinatal outcomes and lack of clear evidence for long-term harm.

Wrap-Up

Use of oral medications for managing GDM is controversial largely because of recent data suggesting harm associated with glyburide and lack of clarity about long-term outcomes in children prenatally exposed to metformin. The ongoing debate reflects differing judgments by experts and clinicians on the strength of the available evidence. We recommend insulin to all our patients with GDM who are not meeting glycemic targets with dietary modification. Most women are interested in oral medication options, so we inform our patients that these medications are worth considering, but the long-term outcomes associated with metformin are still uncertain — and glyburide has potential adverse perinatal outcomes. When women decline insulin, we prefer to offer a trial of metformin over glyburide.