Using GLP-1 Medicines in Women with Obesity Before Pregnancy

Endogenous GLP-1 is produced from proglucagon expressed in the gut, pancreas and brain, which is subsequently cleaved in the intestinal L cells to form glucagon and GLP-1 [35 –37]. The GLP-1 receptor is expressed in the gut, pancreas, brainstem, hypothalamus and vagal afferent nerves [35]. Ingestion of nutrients, especially fat and carbohydrates, stimulates GLP-1 secretion [37]. Activation of the GLP-1 receptor in the gut delays gastric emptying and in the brain increases satiety [35]. Both result in reduced food intake. In the pancreas, GLP-1 receptor activation increases glucose-dependent insulin release and decreases glucagon secretion, facilitating glucose homeostasis [18, 38, 39].

Endogenous GIP (Glucose-dependent Insulinotropic Polypeptide) is synthesised in the K cells of the duodenum [40]. The GIP receptor is expressed on the alpha and beta cells of the pancreas, gut, adipose tissue, brainstem, hypothalamus and vagal afferent nerves. Like GLP-1, GIP receptor activation induces satiety and stimulates insulin secretion in a nutrient-dependent manner. However, GIP enhances glucagon secretion during hypoglycaemia, while GLP-1 decreases glucagon during hyperglycaemia. GIP also directly stimulates lipogenesis, while GLP-1 indirectly promotes lipolysis [41].

Diet-induced weight loss in non-pregnant human adults' results in a reduction in fasting GLP-1 and an increase in GIP [42]. This is presumably adaptive, stimulating increased food intake and storage, and driving weight regain [43]. In pregnancy, there is a progressive decrease in GLP-1 across the pregnancy [44] which likely facilitates increased oral intake and allows shunting of fuel to the growing fetus. The impact of pregnancy on GIP is less clear. Given the rapid increase in GLP-1 production in the peripartum period, it is likely that endogenous GLP-1 has activities beyond appetite regulation and glucose homeostasis, but these are still being elucidated [45].

Incretins are metabolic hormones that stimulate insulin secretion after nutrient intake [46]. Incretin analogues are a family of drugs that mimic the action of endogenous incretin hormones. GLP-1 RAs mimic the action of GLP-1 and act via the GLP-1 receptor, while GIP RAs mimic the action of GIP and act via the GIP receptor [47]. Compared to endogenous incretin hormones which have a very short half-life of a few minutes, exogenous GLP-1 RAs and GIP-RAs have been modified to have a half-life of between 24 h and 7 days depending on the agent [22]. These drugs act in the gut to delay gastric emptying and increase central satiety signalling. They also increase glucose-dependent insulin release and decrease glucagon signalling [47]. GLP-1 RAs may also reduce inflammation both indirectly through weight loss and directly through activation of GLP-1 activation on T cells, which may explain the lower rates of cardiorenal disease in people using these drugs [48].

In non-pregnant adults, large clinical trials of incretin analogues have demonstrated increasing greater efficacy in weight loss and glycaemic control. In 2015, the SCALE study demonstrated that the GLP-1 analogue Liraglutide (3.0 mg weekly) results in 8.0% body weight loss [49], and a mean HbA1c reduction of 1.1% [50] after 56 weeks. In 2021, STEP-1 demonstrated that Semaglutide (2.4 mg weekly) results in 12.4% weight loss over 68 weeks [51], and a mean HbA1c reduction of 1.59% [52]. In 2022 SURMOUNT-1 demonstrated that the GLP-1/GIP analogue Tirzepatide (15 mg weekly) results in 20.9% weight loss over 72 weeks [53] and a mean HbA1c reduction of 1.95% [52]. Although these trials have all included women who inadvertently achieved pregnancy, the trials have specifically excluded women planning pregnancy by requiring women of reproductive age to use two forms of contraception.

GLP-1 RAs are indicated both for weight loss and diabetes control, although prescribing rules vary between countries. While some women may inadvertently become pregnant while using GLP-1 RAs, there is little doubt that others are making a deliberate decision to use these drugs prior to pregnancy [54]. This may be in the context of sub-fertility, such as in women with polycystic ovarian syndrome or anovulation secondary to obesity, or in women who are aiming to lose weight or improve glycaemic control to improve pregnancy outcomes [20, 55]. Importantly, there is no trial evidence to support GLP-1 RAs for this latter indication.

GLP-1 RAs should not be deliberately used in pregnancy, nor would weight loss in pregnancy be desirable [56]. However, given the widespread use of this class of drugs in women planning pregnancy, inadvertent administration in early pregnancy is inevitable. It is presumed that GLP-1RAs do not cross the placenta based on their molecular size. Animal studies regarding this issue are described in the section below. However, use prior to pregnancy or in early pregnancy would be expected to alter the maternal metabolic state and, therefore, both maternal and fetal effects of the drug are possible.

Cesta et al. published an audit of women inadvertently exposed to GLP-1 receptor agonists in the first trimester of pregnancy and reported no increase in the risk of major congenital malformations (adj RR 0.95 (0.72–1.66), and no increase in the risk of major cardiac malformations (adj RR 0.68 (0.42–1.12)) [54]. Dao et al. explored the reproductive safety of GLP-1 based on the databases of six teratology information services based in Europe and Israel. The 168 women exposed were compared to a reference group with diabetes (n = 156) and a reference group with overweight/obesity (n = 163). There was no difference in spontaneous abortion, medical termination of pregnancy and stillbirth [57]. Taken together, these studies demonstrate early reassurance about congenital anomalies and fetal death in the context of short-duration exposure in the first trimester of pregnancy.

In a small retrospective study of 188 pregnancies, mean gestational weight gain was greater in women exposed to GLP-1 prior to pregnancy than unexposed women (11.3 ± 5.4 vs 8.6 ± 7.8 kg, 95% CI 0.60–11.1) [58]. This suggests that pre-pregnancy GLP-1 RA use is associated with an additional ~ 3 kg gestational weight gain. However, previously reported studies using both lifestyle modification and a Very Low Energy Diet (VLED) report gestational weight gain of ~ 3 kg after pre-pregnancy weight loss [59]. Therefore, it is likely that weight loss, and not the drug per se, is driving the additional gestational weight gain. This is further supported by data from Purcell et al. which demonstrates that the rate of weight loss does not impact the rate of weight regain [60].

There are no human studies of GLP-1 RA in lactation. However, transfer to human milk is likely to be low given that GLP-1 RAs are large peptide molecules [61]. Furthermore, GLP-1 ingested by human infants is likely to be largely digested in the gastrointestinal tract of the infant [23].

Muller et al. published a systematic review of animal studies using GLP-1 RAs in pregnancy [23]. This review included data from the European Medicines Agency (EMA) and Federal Drug Administration (FDA) including product information, the Netherlands Pharmacovigilance Centre and the Teratology Information Service of Switzerland. In rat, mouse and rabbit models, GLP-1 RA administration caused retardation of ossification, skeletal variations, major skeletal anomalies and visceral congenital anomalies in a dose dependent manner [23]. These effects were mostly accompanied by decreased maternal food consumption and reduced maternal weight. Although animal models of malnourished mothers have described delayed ossification and skeletal effects, these malnutrition models do not report visceral anomalies. It is also uncertain if the rates of visceral anomalies are higher than that seen in the population of unexposed animals.

While low molecular weight molecules (< 600 Daltons) cross the placenta freely, larger molecules including GLP-RAs (4000–63,000) are presumed not to cross the placenta based on their molecular size [61]. Extendin-4 is a long-acting GLP-1 analogue that is often used in animal studies [62]. One animal study using Exendin-4 found that when mice were injected with the drug and were sacrificed one day later, there was no drug on the fetal side of the placenta. However, when pregnant mice with systemic inflammation were injected with Exendin-4, there was evidence of the drug on the fetal side of the placenta [63]. This suggests while most of the impact of these drugs in pregnancy may be modulated by on the maternal side of the placenta, direct fetal effects cannot be excluded.

When administered in the second trimester of pregnancy in an animal model, Qiao et al. demonstrated that Semaglutide decreases fetal body weight and placental capillary density [44]. Although placental diameter is not affected, there is reduced vessel branching which results in reduced surface areas for nutrient exchange and reduced blood flow [44]. This placental effect likely explains the increase in fetal growth restriction observed in the animal studies.

It is likely that endogenous GLP-1 has effects on fetal development beyond appetite regulation and fetal growth [48, 64]. GLP-1 receptor appears to be rapidly upregulated in the peripartum period, which correlates with increases in surfactant production and early lung function [45]. The impact of exogenous GLP-1 RA administration on these late pregnancy effects is essentially unknown.

Post-partum, when GLP-1 RAs are administered to animals during lactation, the concentration of exenatide in milk in mice, when compared with the maternal concentrations, were < 2.5% for exenatide, 8.3–33% for Semaglutide and 50% for Liraglutide [23].

The evidence that obesity and sub-optimally controlled diabetes adversely affects pregnancy outcomes is over-whelming [65, 66]. However, it is uncertain if the metabolic benefits of GLP-1 RAs used prior to pregnancy are sustained in pregnancy or if these benefits lead to improved pregnancy outcomes. For women with obesity, it is uncertain how much weight loss may be required to improve pregnancy outcomes, and it is uncertain how best to achieve this weight loss while maintaining maternal nutrition.

For women who use GLP-1 RA prior to conception, the ideal time to cease these drugs is unknown. Based on pharmacokinetic studies, women are routinely advised to cease GLP-1 RA's five half-lives before conception (i.e. the half-life of Semaglutide is 5–7 days, so the drug should be ceased ~ 6 weeks before conception) [22, 34]. However, spontaneous conception is often unplanned [67] and choosing the timing for drug cessation may be difficult [68, 69]. Therefore, understanding the implications of early pregnancy exposure is critical [70].

After pre-conception weight loss, maternal weight, body composition and appetite profile are altered. The data regarding the benefits of pre-pregnancy weight loss are sparse [59]. In addition to determining the optimal amount of weight loss, the impact of weight loss on lean tissue in women planning pregnancy must be determined [71, 72]. Data from bariatric surgery, which achieves weight loss by alteration of appetite hormones including GLP-1, suggest that significant weight loss may benefit the mother but be detrimental for the offspring with an increased risk of small for gestational age offspring and preterm delivery [73, 74]. The long-term implications of iatrogenic prenatal GLP-1 receptor activation are also unknown [75].

There is little data to describe what is driving the huge demand for GLP-1 RAs in women planning pregnancy [54, 76]. It is possible that women with obesity presume pre-pregnancy weight loss will reduce the risk of adverse pregnancy outcomes. Alternatively, the use of GLP-1 RAs in women of reproductive age could simply reflect the rising rates of obesity in this demographic. Future research must focus on two areas. First, we must determine the direct effects of GLP-1 RAs on women planning a pregnancy with an ongoing collection of teratogenicity data in women who inadvertently used GLP-1 RAs during pregnancy. Second, we must determine the indirect effects of GLP-1 RAs with a specific focus on the implications of pre-pregnancy weight loss on pregnancy outcomes including maternal and neonatal body composition [71] and appetite regulation [77].

There is compelling evidence that metabolic diseases, including obesity, can adversely impact pregnancy outcomes. GLP-1 RA are being widely used in the preconception period, despite a lack of evidence to demonstrate improved pregnancy outcomes. Human studies of inadvertent exposure to GLP-1 RAs in early pregnancy are reassuring, demonstrating no increase in the risk of congenital abnormalities. However, animal studies suggest that drug use could have a deleterious effect on offspring that is primarily mediated by its effect on maternal nutrition. On this basis, women are advised to avoid pregnancy until five half-lives of the drug have elapsed to allow adequate drug wash-out, and to advocate for trials that specifically include women planning pregnancy.