As Dr. Landon sees it, research published in 2015 put a damper on the use of glyburide, which “had become the number one agent” after an earlier, seminal trial, led by Oded Langer, MD, had shown equivalent glycemic control in about 400 women with GDM who were randomized to receive either insulin or glyburide (N Engl J Med. 2000;343;1134-8). The trial was not powered to evaluate other outcomes, but there were no significant differences in neonatal complications, Dr. Landon said.
One of the 2015 studies – a large, retrospective, population-based study of more than 9,000 women with GDM treated with glyburide or insulin – showed a higher risk of admission to the neonatal intensive care unit (relative risk, 1.41), hypoglycemia in the newborn (RR, 1.40), and large-for-gestational age (RR, 1.43) with glyburide, compared with insulin (JAMA Pediatr. 2015;169[5]:452-8).
A meta-analysis of glyburide, metformin, and insulin showed significant differences between glyburide and insulin in birth weight, macrosomia (RR, 2.62), and neonatal hypoglycemia (RR, 2.04; BMJ. 2015;350;h102). However, “this was basically a conglomeration of studies with about 50 [individuals] in each arm, and in which entry criteria for the diagnosis of GDM were rather heterogeneous,” said Dr. Landon. “There are real problems with this and other meta-analyses.”
The authors of a 2018 multicenter, noninferiority, randomized, controlled trial of about 900 women concluded that their study failed to show that the use of glyburide, compared with insulin, does not result in a greater frequency of perinatal complications. The authors also wrote, however, that the “increase in perinatal complications [with glyburide] may be no more than 10.5%, compared with insulin” (JAMA. 2018;319[17]:1773-80).
That increase, Dr. Landon said, was “not an absolute 10%, but 10% of the complication rate, which probably translates to about 2%.” The only component of a composite outcome (including macrosomia, hypoglycemia, and hyperbilirubinemia) that was significantly different, he noted, was hypoglycemia, which affected 12.2% of neonates in the glyburide group and 7.2% in the insulin group.
Glyburide’s role may well be substantiated in the future, Dr. Landon said during a discussion period at the meeting, through research underway at the University of Pittsburgh aimed at tailoring treatment to the underlying pathophysiology of a patient’s GDM.
The MATCh-GDM study (Metabolic Analysis for Treatment Choice in GDM) is randomizing women to receive usual, unmatched treatment or treatment matched to GDM mechanism – metformin for predominant insulin resistance, glyburide or insulin for predominant insulin secretion defects, and one of the three for combined mechanisms. The study’s principal investigator, Maisa Feghali, MD, of the department of obstetrics, gynecology, and reproductive sciences at the University of Pittsburgh, stressed in a presentation on the study that GDM is a heterogeneous condition and that research is needed to understand the impact of GDM subtypes on treatment response.
Metformin outcomes
Concerns about the impact of metformin on short-term perinatal outcomes focus on preterm birth, Dr. Landon said. The only study to date that has shown an increased rate of prematurity, however, is the “seminal” Metformin in Gestational Diabetes (MiG) trial led by Janet A. Rowan, MBChB, that randomized 751 women with GDM in Australia and New Zealand to treatment with metformin or insulin. The researchers found no significant differences between a composite of neonatal complications but did establish that severe hypoglycemia was less common in the metformin group and preterm birth was more common (N Engl J Med. 2008;358:2003-15).