Key takeaways:

• The ATA published its first updated guideline on thyroid disease in pregnancy since 2017.
• One large change involves no longer treating euthyroid women with levothyroxine, regardless of thyroid autoimmunity.

An updated guideline proposed several shifts in how healthcare professionals manage thyroid health for pregnant women.

In May, the American Thyroid Association published its newest guideline for thyroid disease in preconception, pregnancy and postpartum in the journal Thyroid. The first update of the guideline since 2017 recommends changes in care and an increased emphasis on shared decision-making between patients and healthcare professionals, according to Angela M. Leung, MD, MSc, professor of medicine at the UCLA David Geffen School of Medicine and co-first author on the guideline.

Image: Isaac Martinez, UCLA Endocrine Center

“More discussion is advised with the patient, in conjunction with other considerations such as her age, her fertility status and previous relevant [medical] history, that might be helpful in guiding the decision to undergo treatment or not,” Leung told Healio | Endocrine Today.

One of the biggest changes in the guideline involves whether to treat euthyroid women with levothyroxine. The previous iteration recommended considering treating euthyroid women with levothyroxine if they were positive for thyroperoxidase or thyroglobulin antibodies, particularly in the setting of infertility, assisted reproduction or previous pregnancy loss. The new guideline recommends against levothyroxine treatment of euthyroid women solely on the basis of thyroid autoimmunity.

Elizabeth N. Pearce

“Newer data don’t suggest that there is a huge difference conferred by thyroperoxidase antibody positivity in terms of pregnancy outcomes when somebody is subclinically hypothyroid,” Elizabeth N. Pearce, MD, MSc, professor of medicine at Boston University Chobanian & Avedisian School of Medicine, a Healio | Endocrine Today Editorial Board Member and a coauthor on the guideline, said in an interview. “Based on that, we’re no longer using thyroperoxidase antibody positivity in somebody who’s already subclinically hypothyroid [as] a reason to treat.”

Caroline T. Nguyen, MD, an adjunct associate professor of clinical medicine, obstetrics and gynecology (voluntary) in the division of endocrinology, diabetes and metabolism at the Keck School of Medicine of USC and a coauthor on the guideline, said another change centers on continued antithyroid drug treatment for pregnant women with Graves’ disease, specifically making sure they are taking propylthiouracil instead of methimazole before pregnancy.

“In the past, we used to feel more comfortable with letting patients continue methimazole while they’re not pregnant, and then when they became pregnant, we would switch them to propylthiouracil in the first trimester,” Nguyen told Healio | Endocrine Today. “Subsequent data have brought awareness that perhaps even though we are switching patients early in the pregnancy, the prevalence of congenital malformations is comparable to those who stay on methimazole alone.”

The guideline was endorsed by a multidisciplinary group of organizations that includes obstetrician and gynecologist associations. Rachel Pessah-Pollack, MD, FACE, clinical professor of medicine in the Holman Division of Endocrinology, Diabetes & Metabolism at NYU Langone Health and a Healio | Endocrine Today Editorial Board Member, said educating obstetricians and primary care professionals about the new recommendations will be crucial for implementation.

“I have many calls with obstetricians locally where I’ll get them on the phone and I’ll say, ‘Did you see the most recent ATA thyroid in pregnancy guideline?’” Pessah-Pollack, who was not an author on the guideline, told Healio | Endocrine Today. “I think having commentary in one of the obstetrician journals would be important to help with the implementation of these recommendations.”

Levothyroxine recommendations

One of the biggest shifts in the 2026 guideline centers on thyroid autoimmunity and levothyroxine. A new recommendation states that women who are euthyroid should not receive levothyroxine, even if they are positive for thyroperoxidase antibodies.

The new guidance was based on findings from three randomized controlled trials. In a phase 3 trial published in The Lancet Diabetes & Endocrinology in 2022, levothyroxine did not increase the likelihood for live births compared with placebo among euthyroid women who were positive for thyroperoxidase antibodies and had previously experienced recurrent pregnancy loss. The results were the same as another trial published in The New England Journal of Medicine in 2019.

A randomized clinical trial published in JAMA in 2017 found women who were positive for thyroperoxidase antibodies and underwent IVF did not have improved pregnancy outcomes with levothyroxine compared with those who did not receive the medication.

Nguyen said the new recommendations align closer with guidance provided by the American College of Obstetricians and Gynecologists (ACOG) and American Society for Reproductive Medicine.

“In 2017, the guideline said, ‘consider treating,’ and now it says, ‘don’t treat,’” Nguyen said. “That is going to be a big shift for a lot of clinicians who have been accustomed to treating.”

Rachel Pessah-Pollack

Antibody status should not be used to determine levothyroxine treatment for women with subclinical hypothyroidism, according to the recommendations, but there may be benefit if a woman is treated very early in pregnancy, Pearce said.

“We now have good trials suggesting levothyroxine started in or after the second trimester does not improve adverse pregnancy or child neurocognitive outcomes,” Pearce said. “It’s biologically plausible that there might be a benefit of intervention earlier in pregnancy if it’s caught in the first trimester and not later in gestation where we know it’s probably not going to make a difference.”

Pessah-Pollack said she agrees with the change and believes that the new recommendations on levothyroxine decision-making will have a positive effect on clinical care.

“Patients get concerned when they see that those antibodies are positive, thyroid levels are normal and are looking to get pregnant and wonder what the significance of it is,” Pessah-Pollack said. “Having these newer data will provide guidance for our patients and avoid unnecessary levothyroxine treatment in this patient population.”

Antithyroid drugs in pregnancy

The guideline also highlights the use of antithyroid drug therapy for women with Graves’ disease.

Leung said the recommendations in the 2026 guideline are not drastically different from 2017, but details have been added about the risks of antithyroid drug use before and during pregnancy. The guideline states that antithyroid drug use during pregnancy may raise risk with congenital anomalies, fetal goiter and fetal or neonatal thyroid dysfunction.

“We provide the absolute risks of potentially developing a congenital malformation or embryopathy related to antithyroid drugs,” Leung said. “That number is up to 3% additional risk above baseline for propylthiouracil, and then additional up to 5% for methimazole and carbimazole. This gives the clinician and the patient themselves a little bit more discrete and concrete perception of the absolute risk.”

When considering antithyroid drug therapy, shared decision-making should be used during preconception, the guideline states. If a woman is at low risk for disease relapse, therapy should be stopped in first trimester of pregnancy. Women who are planning pregnancy and need antithyroid therapy should be switched from methimazole to propylthiouracil in the preconception period, while those who conceive while already taking methimazole should have a discussion of the pros and cons of switching during what may be critical periods of fetal organ development, according to the guideline.

As such, Nguyen said it is preferrable for women to use propylthiouracil before pregnancy to avoid switching treatment during the first trimester.

“One of the major changes in this guideline was this recommendation that patients who are interested in pregnancy and trying for pregnancy actively should be switched to propylthiouracil in the preconception period to avoid having to have this exposure to both antithyroid drugs once pregnant,” Nguyen said.

For women who conceive while using methimazole, staying on that drug can be considered rather than switching to propylthiouracil, depending on the patient’s gestational week and antithyroid requirements, especially if propylthiouracil is not available, Nguyen said. Data published in The Journal of Clinical Endocrinology & Metabolism in 2019 showed that a higher percentage of women who received both propylthiouracil and methimazole had babies with congenital malformations within 2 years of birth compared with women receiving methimazole alone.

“Because of these new data, it raises the question of whether patients should be switched or not,” Nguyen said. “If the prevalence of malformations is still high, it becomes more nuanced. Things that I take into consideration are how early [is the patient] in first trimester. If we’re very early before organogenesis, then I think switching makes a lot of sense because then you can still remove the methimazole and we know patients are most sensitive between gestational weeks 6 to 10. If they’re coming in at week 11, 12 or 13, then the utility of switching them to propylthiouracil at that point may not be as helpful because we’re past the point of increased risk of congenital malformations.”

Pessah-Pollack said the recommendations regarding whether to use propylthiouracil or methimazole at 16 weeks’ gestation or later stood out to her. The guideline states that after 16 weeks, “it remains unclear” whether patients should continue propylthiouracil or switch to methimazole.

“We have this evidence gap, and we need future studies to help us determine if there is a preferred treatment option for hyperthyroidism at 16 weeks’ gestation. This gives us options in terms of counseling our patients and tailoring the best treatment based on how they are doing on their current regimen and reviewing the risks and benefits,” Pessah-Pollack said. “If you have a patient that is euthyroid on the lowest dose of propylthiouracil, you don’t have to switch them over to methimazole because you hit that 16-week point.”

Thyroid function testing

The updated guideline lists 12 risk factors for thyroid dysfunction screening, which is a change from the 17 risk factors listed in the 2017 guideline.

Caroline T. Nguyen

“The list is now a shorter, trimmer list,” Leung said. “That is hopefully going to help streamline who would benefit from thyroid function testing in pregnancy.”

Pearce said the previous list of risk factors was “overly broad,” and the authors opted to remove weaker predictors for thyroid dysfunction such as age and BMI.

“When you take those things off the list, you cut down substantially on the number of people who need testing,” Pearce said. “We also tried with this guideline to be a little more intentional about how often to test.”

The guideline also discusses additional options for assessing thyroid function in pregnancy. Leung said TSH measurements are physiologically decreased by human chorionic gonadotropin during pregnancy, particularly in early pregnancy. Because of this, she said, healthcare professionals need to interpret both TSH and thyroxine levels in the context of pregnancy when assessing thyroid function. With different methods of measuring T4 available, the guideline details advantages and disadvantages of various testing methods and recommends free T4 as a complement to TSH over other measures, Leung said.

“The uncertainties around the reliability of free T4 seem to have improved in recent years as such that this is going to be the preferred standard to complement TSH,” Leung said.

Moving forward

Despite the emergence of several key trials since 2017, there are still gaps in research on managing thyroid disease during pregnancy, Leung said. A section near the end of the guideline highlights 13 of the largest areas where more studies are needed.

Leung said one of the biggest topics in need of more research is how much women with mild or moderate iodine deficiency are at risk for complications during pregnancy.

“We have a lot of data on severe iodine deficiency, but it’s more the mild and moderate types, which affects a lot more women,” Leung said.

Pearce said there are still many recommendations in the guideline that are based on moderate- or low-quality evidence. Pearce said one of the biggest research gaps centers on determining antithyroid drug dosing for pregnant women with Graves’ disease.

“We don’t have good prospective data for how best to dose antithyroid drug in pregnancy,” Pearce said. “What is the optimal treatment target? We make recommendations that high normal to slightly elevated free T4 is probably the safest, but we don’t know exactly where that should be to optimize fetal thyroid hormone availability.”

A goal for the guideline authors is to harmonize the new recommendations in both the endocrinology and obstetrics fields. In addition to being endorsed by endocrinology organizations, the guideline received endorsements from ACOG, the European Board & College of Obstetrics and Gynaecology and the European Society of Human Reproduction and Embryology.

Pessah-Pollack said endorsements by obstetrics organizations are helpful, and there will be a need for wider educational tools to inform healthcare professionals.

“We are going to need an educational outreach and collaboration with the obstetrics and primary care communities to ensure that we’re all similarly minded regarding whether or not we need to be treating,” Pessah-Pollack said.

• For more information:
• Angela M. Leung, MD, MSc, can be reached at amleung@mednet.ucla.edu.
• Caroline T. Nguyen, MD, can be reached at endocrinology@healio.com.
• Elizabeth N. Pearce, MD, MSc, can be reached at elizabeth.pearce@bmc.org.
• Rachel Pessah-Pollack, MD, FACE, can be reached at rachel.pessah-pollack@nyulangone.org; or on Instagram @drpessahpollack.