Pregnancy outcomes following first trimester exposure to semaglutide

Abstract

Background and objective

To review pregnancy outcomes in women exposed to semaglutide during early pregnancy.

Methods

A retrospective review of an obstetric database at a tertiary obstetric hospital in Brisbane, Australia.

Results

Thirteen women were exposed to semaglutide in first trimester. One infant had significant cardiac anomalies in the setting of very poor maternal glycaemic control in first trimester, as well as maternal obesity and hypertension.

Conclusion

This study adds to the growing literature suggesting there is no increased risk of major congenital anomalies following first trimester exposure to semaglutide, however, until larger studies are available to confirm these findings, reproductive-age women using semaglutide should practice effective contraceptive methods. Preconception counselling of all women with diabetes mellitus is critical to avoid adverse pregnancy outcomes.

Background and objective

Semaglutide is a glucagon-like peptide-1 receptor agonist (GLP-1 RA) administered weekly subcutaneously, which was listed on the Australian Pharmaceutical Benefits Scheme (PBS) in July 2020 for the management of type 2 diabetes mellitus (T2DM). Two other GLP-1 RA, exenatide and dulaglutide, had been previously listed on the PBS in August 2010 and June 2018 respectively. The primary glucose-lowering mechanism of GLP-1 RA is stimulation of glucose-dependent insulin release from the pancreatic islets. ¹ Additional mechanisms of action include inhibition of inappropriate glucagon secretion, suppression of hepatic gluconeogenesis, and delayed gastric emptying resulting in early satiety and weight loss. ¹ In addition to improving glycaemic control and weight loss, the use of semaglutide in individuals with T2DM is associated with reduction in the risk for major cardiovascular events, cardiovascular mortality in high-risk individuals, fatal or non-fatal ischaemic stroke, and progression of chronic kidney disease. ² The prescription of a GLP-1 RA within 24 months preceding a pregnancy was associated with reduced risks of hypertensive disorders of pregnancy, preterm delivery, and caesarean delivery in a study population, 38% of whom had T2DM. ³

An oral form of semaglutide was approved for medical use by the Therapeutic Goods Administration in February 2022.

An observational population-based cohort study of 3,514,865 women in the United States, four Nordic countries, and Israel, between 2009 and 2020, suggests a rapidly increasing number of women are conceiving while taking GLP-1 RA for the management of T2DM. ⁴ Review of the U.S. MarketScan database of commercially insured women with T2DM found the periconception exposure to GLP-1 RA increased from approximately 50 to 200 per 100,000 pregnancies between 2012 and 2021. ⁴ The Australian Institute of Health and Welfare reported that the prevalence of T2DM in Australian women less than 40 years of age, 40–44 years of age, and 45–49 years of age were 0.3%, 1.9% and 3.1% respectively in 2021. ⁵ In addition, an increasing number of reproductive-aged women may be taking GLP-1 RA for weight reduction. In the decade between 2008 and 2018, the prevalence of obesity in Australian women aged 25–34 years rose by 110%, and in women aged 35–44 years by 56%.^6,7 In 2017, 21% of Australian women were obese at the time of their first antenatal visit. ⁸

Maternal exposure to GLP-1 RA other than exendin-4 in animals has been associated with reduced fetal growth, retardation of ossification and skeletal variants in a dose-dependent fashion. ⁹ The TGA lists semaglutide as a pregnancy category D medication. The Best Use of Medicines in Pregnancy database states that GLP-1 RA should not be used in pregnancy as there is inadequate safety information regarding safety for the fetus. ¹⁰ Effective contraception is recommended for any woman of reproductive age taking GLP-1 RA. Due to the long half- life, current guidelines recommend cessation of semaglutide at least eight weeks prior to conception. ¹¹ Semaglutide was not detectable in the breast milk of mothers taking the medication subcutaneously, and their breastfed infants experienced no adverse effects, thus no precautions appear to be necessary in women breastfeeding while being treated with semaglutide. ¹² A scoping review reported an absence of primary data about the role of GLP-1 RA on the fertility and reproductive health outcomes of women without polycystic ovarian syndrome. ¹³

The objective of this study was to describe pregnancy outcomes following maternal exposure to semaglutide in first trimester.

Methods

A retrospective review of an obstetric database of women booked for antenatal care with a diagnosis of pre-conception T2DM at the Mater Mothers Public Hospital, Brisbane, Australia, a tertiary referral obstetric hospital, who had an estimated date of delivery between January 2021 and December 2023, was reviewed for medication exposures, pregnancy outcomes, and whether the women recalled having preconception counselling. Approximately 6100 deliveries occur at the Mater Mothers Public Hospital per year. Medication exposures, and patient recall of preconception counselling, were based on review of electronic medical records of specific history taken at each woman's first visit to the obstetric medicine service.

Results

Of the 18,354 women who had an estimated date of delivery in the calendar years 2021–2023, 150 (0.82%) had been diagnosed with T2DM prior to conception. Twelve of these 150 women with pre-existing T2DM (8%) were exposed to semaglutide in first trimester (Table 1). The data from an additional woman who was taking semaglutide for weight reduction, identified through preconception care for other conditions, was also included. Five women were nulliparous, mean age was 33 years, nine of the 12 women with pre-existing T2DM had BMI greater than 30 kg/m², and six of 11 had an HbA1c greater than 6.5% (> 48 mmol/mol).

Table 1.

Pregnancy outcomes with first trimester exposure to semaglutide.

Age	P	BMI	Indication	Rx precon	HbA1c (%)	Last expose	Rx preg	Gest. deliver	Bthwt	Mat. Cx	Fetal/neonatal Cx	Other maternal conditions
33	2	35	T2 DM	M	9	4	I	38.3	3910	nil	nil
35	0	37	T2DM	M	12.5	4	M	38.1	3288	nil	Cardiac anomaly	Hypertension, IIH, smoker
37	3	43	T2 DM	M	5.9	6	M,I	38.4	3484	nil	nil	Hypertension
31	5	40	T2DM	I	5.6	7	I	35.6	4800	nil	Macrosomia, polyhydramnios, RDS	Addison, OSA, previous NND
35	2	30	T2 DM	M,I	9	6	M,I	31.6	1035	PET	FGR
31	2	34	T2DM	nil	NA	6	I	38.2	3950	nil	nil	PCOS
35	0	42	T2DM	M,I	9.6	6	M,I	36.4	3360	PET	nil
39	2	29	T2DM	M	5.1	6	M,I	34.5	3896	nil	Fetal distress, RDS
27	1	29	T2DM	M	4.8	4	M,I	36.5	3080	nil	nil
26	0	26	T2DM	M	6.2	2	M,I	38.1	3348	nil	nil
34	0	40	T2DM	M	7.9	6	M,I	40	3462	nil	nil
33	0	30	T2DM	M, E	7.5	6	M,I	38.3	3852	nil	nil
35	1	NS	Wt loss	Nil	NS	6	nil	38.5	3930	nil	nil

P: parity; BMI: body mass index (kg/m²); Rx precon: treatment preconception in addition to semaglutide; Last expose: gestation at which last dose semaglutide received (weeks); Rx preg: treatment during pregnancy; Gest deliver: gestation delivery (weeks); Bthwt: neonatal birthweight (g); Mat. Cx: maternal complications; fetal/neonatal Cx: fetal/neonatal complications; M: metformin, I: insulin; E: empagliflozin; NS: not stated; HbA1c: HbA1c at first visit; RDS: respiratory distress syndrome; FGR: fetal growth restriction; IIH: idiopathic intracranial hypertension; OSA: obstructive sleep apnoea; PCOS: polycystic ovarian syndrome; NND: neonatal death; T2DM: type 2 diabetes mellitus.

Of the 12 women with T2DM taking semaglutide in early pregnancy, nine, three and one were also being treated with metformin, insulin and empagliflozin respectively. The woman taking empagliflozin ceased this at six weeks gestation. None of the women recalled having received preconception counselling.

In terms of adverse outcomes, one fetus had significant cardiac anomalies in the setting of the mother having T2DM with early pregnancy HbA1c of 12.5% (113 mmol/mol), body mass index of 37 kg/m², and co-existent hypertension. There were no other fetal or neonatal abnormalities. One infant had fetal growth restriction in the setting of maternal preeclampsia. Five infants were delivered prior to 38 weeks gestation, the indications being preeclampsia (2), fetal distress (1), macrosomia with polyhydramnios (1), and pre-term labour (1). Three neonates required treatment for hypoglycaemia.

Discussion

With the rising incidence of both obesity and T2DM, and consequent increase in use of GLP-1 RA in women of childbearing age, the safety of GLP-1 RA in pregnancy is particularly relevant. Ex-vivo and in vivo studies have shown negligible placental transfer of GLP-1 RA.^14,15 Six previous case reports have described pregnancy outcomes following maternal exposure to GLP1-RA during pregnancy, one with semaglutide^14,16–18 (Table 2). No neonatal malformations were described. Notably, one woman was treated with semaglutide for her entire pregnancy.

Table 2.

Previous case reports of GLP-1 RA use in pregnancy.

Author	Medication	Last maternal dose (weeks)	Neonatal outcome
Greco et al.	Liraglutide	13	Healthy at term
Ivansevic et al.	Liraglutide	Term	Healthy at term Well at age 2 years
Burlina et al.	Dulaglutide	15	Neonatal jaundice Otherwise healthy
Alghamdi et al.	Dulaglutide	13	Healthy at term
Williams et al.	Exenatide	14	Healthy at term
Skov et al.	Semaglutide	3	Normal development at 6 months age

Two multicentre, observational, prospective cohort studies have suggested maternal exposure to GLP-1 RA is not associated with fetal malformations.^14,21

Cesta et al. performed an observational population-based cohort study using data from four Nordic countries, the U.S. MarketScan Database, and the Israeli Maccabi Health Services database, assessing the risk of major congenital malformations (MCM) in pregnancies with maternal periconception exposure to second-line noninsulin antidiabetic medication. ⁴ Periconception exposure was defined based on the filling of prescriptions of the respective drug class from 90 days before the first day of the last menstrual period to the end of the first trimester. Of the women, 938 (6.2%) had periconception exposure to GLP-1 RA. Periconception exposure to GLP-1 RA was associated with a prevalence of MCM of 8.3%, compared with 5.3% in all infants born to mothers with T2DM, and 3.7% in infants born to mothers in the general population. The higher rate of congenital malformation with GLP-1 RA could be related to a higher HbA1c in this group compared to all patients with T2DM, although the HbA1c data in this study was incomplete. Only the United States and Israeli cohorts had periconception HbA1c data, which found that HbA1c levels were highest among those using insulin or second-line non-insulin anti-diabetic medication, compared to other pregnant women with T2DM treated with metformin or diet alone. Compared with insulin, the adjusted relative risk for MCMs with GLP-1 RA was 0.95 (95% CI [0.72–1.26]), consistent with no substantial increased risk for MCMs among infants with periconceptional exposure to GLP-1 RA.

Dao et al. performed a multicentre, observational prospective cohort study based on the database of six Teratology Information Services, comparing pregnancy outcomes in women exposed to GLP-1 RA in first trimester for the management of T2DM (n = 46) or obesity (n = 117). ²¹ Reference groups of 156 pregnancies of women with T2DM managed with non-GLP-1 RA therapy in first trimester (predominantly metformin), and 163 pregnancies of women who were overweight/obese without T2DM were used for comparison. Of the women with periconception exposure to GLP-1 RA, 51 (30%) were prescribed semaglutide. GLP-1 RA was initiated pre-conception in 88% of the women in the exposed group and stopped at a median gestational age of 5 weeks. Exposure to GLP-1 RA in first trimester was not associated with increased risk of MCM compared with T2DM managed with non-GLP-1 RA therapy, or women who were obese without T2DM (2.6% vs 2.3% vs 3.9%). Cox proportional cause-specific hazard models indicated no increased risk of pregnancy losses in the GLP-1 RA versus the T2DM managed with non-GLP-1 RA therapy and the overweight/obese reference groups.

A major concern from our case series was that none of the women recalled having received preconception counselling, possibly reflected in the poor glycaemic control periconception, and pregnancy occurring while taking GLP-1 RA therapy. Preconception counselling is associated with better glycaemic control three months preconception and in the first trimester, and reduced risk of adverse pregnancy outcomes. ²¹ The prevalence of self-reported recall of preconception counselling by health providers is lower in women with T2DM than T1DM.^22–25 Distribution of a preconception counselling DVD by diabetes care teams and general practices reported the DVD was viewed by 53% and 20% of women with T1DM and T2DM respectively. ²⁶ Effective preconception care (defined as HbA1c less than 6.5% at conception) occurred in only four of 97 pregnancies (4.1%) in women with T2DM at a University hospital in France. ²⁷

There were several limitations to this study. Early pregnancy loss was not assessed. In Australia the supply of GLP-1 RA was restricted during the time of the study, and a significantly higher proportion of reproductive-aged women may be treated with semaglutide as availability improves. Some women may not have divulged their use of semaglutide. The women's primary care health records were not reviewed regarding pre-conception counselling, which the women may have received, and not recalled.

Conclusion

In conclusion, based on the limited data available, it appears unlikely that exposure to GLP1-RA in first trimester is associated with increased risk of major congenital malformations. This may be reassuring to women with unplanned pregnancies while taking this class of medication. Larger studies will be useful to validate this finding. Pre-conceptional counselling and the achievement of optimal glycaemic control periconception is critical to minimise the risk of adverse pregnancy outcomes in women with T2DM.