Conflicting findings, guidelines support cautious approach to mild thyroid dysfunction in pregnancy
Subclinical hypothyroidism is a common diagnosis, present in approximately 3% to 4% of reproductive-age women. Although a mild form of hypothyroidism — defined as elevated thyroid-stimulating hormone with normal free thyroxine levels — the condition has been associated with infertility, increased risk for adverse pregnancy and neonatal outcomes, and possibly with increased risk for neurocognitive deficits in offspring.
“Maternal thyroid hormones are critical for the normal neurologic development of the child as well as for the maintenance of the pregnancy,” Alex Stagnaro-Green, MD, MHPE, MHA, regional dean and interim head of the department of family and community medicine and professor of medicine, obstetrics and gynecology and medical education at the University of Illinois College of Medicine Rockford, told Endocrine Today. “Severe, overt cases of hypothyroidism have been associated with miscarriage and preterm delivery. However, even subclinical hypothyroidism has been associated with miscarriage, preterm delivery, gestational hypertension, preeclampsia, and decreased IQ in the child. There are important sequalae of having an underactive thyroid during pregnancy for both the mother and the fetus.”
Despite well-established recommendations for the treatment of overt hypothyroidism among pregnant women, no consensus has been reached on whether to treat women with subclinical disease. Important uncertainties remain regarding the screening and management of maternal thyroid status in optimizing perinatal outcomes. Researchers continue to debate the very definition of subclinical hypothyroidism, which can differ between nonpregnant or pregnant states.
“The controversy is not when you have very bad thyroid function in pregnancy, which can affect mental development in the baby, but borderline, so-called — badly called — ‘subclinical’ thyroid function, better called ‘mild thyroid dysfunction,’” Terry F. Davies, MBBS, MD, FRCP, FACE, the Florence and Theodore Baumritter professor of medicine at the Icahn School of Medicine at Mount Sinai and Mount Sinai Beth Israel Medical Center, New York, told Endocrine Today. “There is no doubt that women with thyroid antibody positivity and a normal TSH also have reduced thyroid hormone response to pregnancy. Women with a slightly increased serum TSH are the next level. It’s been controversial as to whether this elevated TSH has any effect on the baby’s mental development. Studies suggesting that it does have not been very convincing, and some larger studies couldn’t show an effect.”
In an updated guideline for the diagnosis and management of thyroid disease during pregnancy issued by the American Thyroid Association in 2017, researchers noted that subclinical disease is associated with adverse pregnancy outcomes in some, but not all studies, partly because studies use different cutoffs to define elevated TSH concentration, whereas others do not account for thyroid peroxidase (TPO) antibody status. However, only a small number of studies have investigated the impact of levothyroxine therapy on pregnancy complications in women with subclinical disease. That absence of clear evidence has led to debate about whom to treat and when.
“It can be confusing, even for the endocrinologist,” Caroline Nguyen, MD, assistant professor of clinical medicine at Keck School of Medicine of the University of Southern California, told Endocrine Today. “The guidelines are a very good source, and we are getting more papers that are helping to answer these questions.”
Levothyroxine therapy is unanimously recommended during pregnancy for women with overt hypothyroidism, defined as below-normal serum free T4 in conjunction with elevated TSH, or a TSH of at least 10 mIU/L irrespective of free T4 level, Inés Velasco, MD, PhD, a researcher with the pediatrics, gynecology and obstetrics unit at Riotinto Hospital in Huelva, Spain, and colleagues wrote in a 2018 review published in the European Journal of Endocrinology.
However, in cases of subclinical hypothyroidism or autoimmune thyroid disease during pregnancy, guidelines have undergone “frequent modifications” to incorporate the best evidence available during the last decade, the researchers wrote.
“Whether or not women with mild thyroid function test abnormalities, such as subclinical hypothyroidism, should be treated with levothyroxine is debated because there is no conclusive evidence to show that subclinical hypothyroidism is associated with a higher risk for adverse outcomes, or that levothyroxine would lower the risk of adverse outcomes,” Tim Korevaar, MD, PhD, a postdoctoral fellow and clinical epidemiologist at Erasmus University Medical Center in Rotterdam, the Netherlands, told Endocrine Today. “There are no good studies that show that subclinical hypothyroidism before pregnancy affects the chances of becoming pregnant or the risk of a miscarriage, and there is conflicting data on the risk of adverse outcomes in women who develop subclinical hypothyroidism while pregnant.”
The most recent data, Korevaar said, seem to indicate that subclinical hypothyroidism is particularly a risk factor for adverse pregnancy outcomes when it coincides with TPO antibody positivity, a marker for the presence of autoimmune thyroid disease. Additionally, new data suggest that treatment with levothyroxine has risks, including hyperthyroidism symptoms and, in the case of overtreatment, potential risk for suboptimal fetal development.
In a practice bulletin issued in 2015, the American College of Obstetricians and Gynecologists stated there is no evidence that identification and treatment of subclinical hypothyroidism during pregnancy improves outcomes. The most recent clinical practice guidelines issued by the Endocrine Society (2012) and the European Thyroid Association (2014) for managing thyroid dysfunction during pregnancy both endorse levothyroxine replacement in the presence of subclinical disease, independent of the presence of thyroid antibodies; the recommendation level is weaker among women with subclinical hypothyroidism who are TPO antibody-negative.
The 2017 ATA guideline recommends, first, the evaluation of TPO antibody status for pregnant women with TSH concentrations of at least 2.5 mIU/L. Levothyroxine therapy is recommended for women who are TPO antibody-positive with a serum TSH greater than the pregnancy-specific reference range (strong recommendation, moderate quality evidence).
The recommendation reflects a change in the TSH upper limit range, Nguyen said.
“Where it used to be between 2.5 mIU/L and 3 mIU/L, the guideline authors found that to be too aggressive, and it was leading to overtreatment of patients,” Nguyen said. “The 2017 guideline recommends that we use trimester-specific, lab-specific reference ranges for the treatment of pregnant women with thyroid disease.”
The controversy surrounds so-called “gray zones,” where recommendations state the physician can consider levothyroxine treatment. These include patients with TPO antibody positivity and a TSH level between 2.4 mIU/L and 4 mIU/L, and TPO antibody-negative women with a TSH level between 4 mIU/L and 10 mIU/L. The ATA guideline states that levothyroxine “may be considered” for women with TSH concentrations of at least 2.5 mIU/L and below the upper limit of the pregnancy-specific reference range (weak recommendation, moderate quality evidence).
“The reason that was written that way was, while there was some good, observational data that there could be potential benefit for treatment, there wasn’t any solid data that showed that intervention made a huge difference,” Nguyen said. “That was the controversy they left up to the physician to decide.”
Still, some experts maintain that levothyroxine in pregnancy is safe and that physicians should err on the side of treating in most cases of subclinical disease.
“Some people don’t want to treat [subclinical disease]. I don’t find that comfortable,” Davies said. “Until they show me data that levothyroxine has absolutely no effect, I am going to treat with something that is harmless.”
Debate continues over the optimal time to initiate levothyroxine treatment for women with subclinical thyroid dysfunction and whether earlier vs. later treatment makes a difference in maternal and offspring outcomes. In a study published in The New England Journal of Medicine in 2017 assessing the association between subclinical disease and offspring IQ, Brian M. Casey, MD, professor of obstetrics and gynecology at the University of Texas Southwestern Medical Center in Dallas, and colleagues analyzed data from pregnant women with subclinical hypothyroidism (n = 677) and hypothyroxinemia (n = 526) randomly assigned to levothyroxine or placebo during the second trimester until delivery, with thyroid function assessed monthly. Children underwent annual developmental and behavioral testing for 5 years. The primary outcome was IQ score at age 5 years.
In the subclinical hypothyroidism group, the median IQ score of children was 97 in the levothyroxine group vs. 94 in the placebo group (P = .71). In the hypothyroxinemia group, median IQ score was 94 in the levothyroxine group and 91 in the placebo group (P = .3).
Similarly, a randomized controlled trial with nearly 800 women conducted by John H. Lazarus, MD, FRCP, FRCOG, FACE, emeritus professor of clinical endocrinology at Cardiff University School of Medicine in Cardiff, Wales, and colleagues in 2012, demonstrated that antenatal screening (at a median gestational age of 12 weeks) and maternal treatment for hypothyroidism did not result in improved cognitive function in children at age 3 years.
“There are numerous studies showing a strong association between subclinical thyroid disease and adverse pregnancy outcomes, and based on those observational studies came a number of prospective trials, the best known being the Lazarus study and the Casey study, both looking at IQ of the child at ages 3 to 5 years — neither showed an impact of treating vs. not treating,” Stagnaro-Green said. “The problem with both studies is that they began treating in the second trimester, missing the critical time period.”
Rima Dhillon-Smith, PhD, MRCOG, MBChB, a specialist registrar in obstetrics and gynecology and clinical lecturer in early pregnancy and reproductive medicine at the University of Birmingham, United Kingdom, and colleagues sought to answer the question of early intervention with the design of the TABLET trial, a double-blind, placebo-controlled study assessing the benefit of levothyroxine therapy prescribed before and during pregnancy to 952 euthyroid women with thyroid antibody positivity (mean age, 33 years) who reported at least one miscarriage or infertility.
That trial, too, had a disappointing finding: Researchers found that euthyroid women with thyroid antibody positivity prescribed levothyroxine before and during pregnancy were no more likely to experience a live birth at 34 weeks’ gestation than similar women assigned placebo, 37.4% vs. 37.9%, respectively.
The researchers observed no between-group differences in pregnancy loss, preterm birth rates or neonatal outcomes. In a subgroup analysis for the outcome of live birth, there was no difference observed in women with higher TSH values above 2.5 mIU/L, taking levothyroxine therapy vs. placebo, Dhillon-Smith said.
Women in the levothyroxine group experienced a nonstatistically significant greater number of adverse events compared with women assigned placebo (5.9% vs. 3.8%), including a higher rate of preeclampsia and gestational diabetes, Dhillon-Smith said. The findings were presented at the Endocrine Society Annual Meeting in March.
“We believed the mechanism by which levothyroxine may have worked, as suggested in previous smaller studies, was by maintaining a tight thyroid control and preventing progression to thyroid disease in pregnancy and therefore preventing adverse outcomes,” Dhillon-Smith told Endocrine Today. “We found no difference in live birth rate between those women on levothyroxine compared with placebo. We were surprised by the findings of our study, and it has led to us reconsider the underlying process by which thyroid antibodies are causing miscarriage and preterm birth, as this does not appear to be driven through thyroid function.”
Korevaar said risks of overtreating and medicating unnecessarily are arguments against levothyroxine therapy, and the issue remains a difficult topic despite the TABLET findings.
Many clinicians can provide anecdotal evidence of a beneficial effect of levothyroxine treatment, he said.
“Since the course of thyroid function throughout pregnancy is attenuated in a large proportion of TPO antibody-positive women, my advice is to follow up these women and check thyroid function every 3 months until conception, and again when around 6 to 8 weeks pregnant,” Korevaar said.
Some experts, however, have noted a selection bias in the TABLET study — 30% of women in the placebo group experienced pregnancy loss — complicating interpretation of the findings.
“One of the major problems was it didn’t focus on the group shown to have the strongest association with pregnancy loss, which is not the miscarriage or infertile group, but unselected women,” said Stagnaro-Green. “That is the major deficiency in that study — it didn’t look at the group most critical to be analyzed. A study has to be conducted with unselected women with subclinical thyroid disease, plus or minus antibody positivity, screening them preconception. Then, that will truly give us the answer for subclinical thyroid disease, as well as for thyroid autoimmunity.”
In a 2018 review published in the Journal of the Endocrine Society,Spyridoula Maraka, MD, MS, an assistant professor at the University of Arkansas for Medical Sciences, and colleagues wrote that there is insufficient evidence for or against routine levothyroxine therapy to aid conception among TPO antibody-negative women with infertility who are not undergoing assisted reproduction techniques. However, the ATA issued a weak recommendation that administration of levothyroxine “may be considered in this setting” given its ability to prevent progression to overt hypothyroidism once pregnancy is achieved.
“When it comes to preconception, I would tell individuals that the answer is not known,” Stagnaro-Green said. “I would give them potential pros and the potential cons of treating. If it is a woman aged 39 years, who never tried to get pregnant before, is nearing the end of her fertile years and has decreased chances for pregnancy, I would probably recommend she be treated. If she had multiple miscarriages, I would recommend she be treated. If she’s aged 25 years, has never been pregnant and is not excited about taking medicine, I would support not treating that person. It depends on, to a large extent, the patient and the scenario, and that is what always happens when the science isn’t there to give all the answers.”
Nguyen said women who already know they have subclinical disease should plan pregnancies, when possible, with their health care provider. “This requires talking to your physician or your endocrinologist and checking your thyroid status,” Nguyen said. “‘Subclinical’ is based on your TSH and free T4 levels, but what is more important is getting that TPO status. If [TPO antibody] positive, we go through that same dilemma.”
Preconception treatment with levothyroxine therapy is “reasonable” for thyroid antibody-negative women with TSH concentrations of at least 4 mIU/L, said Dhillon-Smith. She noted that careful consideration should be applied when defining subclinical hypothyroidism.
“The big message for patients is, yes, if you have subclinical disease, definitely see your physician before conceiving and get tested so that things could be optimized, or at least you can have a discussion before you get pregnant,” Nguyen said. “Once you get pregnant, you do need to be monitored more closely. The demands in pregnancy go up, and the concern is whether, in the setting of subclinical disease, the thyroid will be able to meet the demand.” – by Regina Schaffer
Click here to read the , "Should all pregnant women be screened for thyroid dysfunction?"
- Alexander EK, et al. Thyroid. 2017;doi:10.1089/thy.2016.0457.
- Casey BM, et al. N Engl J Med. 2017;doi:10.1056/NEJMoa1606205.
- Dhillon-Smith RK, et al. N Engl J Med. 2019;doi:10.1056/NEJMoa1812537.
- Lazarus JH, et al. N Engl J Med. 2012;doi:10.1056/NEJMoa1106104.
- Maraka S, et al. J Endocr Soc. 2018;doi:10.1210/js.2018-00090.
- Shields BM, et al. J Clin Endocrinol Metab. 2013;doi:10.1210/jc.2013-2768.
- Velasco I, et al. Eur J Endocrinol. 2018;doi:10.1530/EJE-17-0598.
- For more information:
- Terry F. Davies, MBBS, MD, FRCP, FACE, can be reached at the Icahn School of Medicine at Mount Sinai, 1 Gustave L. Levy Place, New York, NY 10029; email: firstname.lastname@example.org.
- Rima Dhillon-Smith, PhD, MRCOG, MBChB, can be reached at the University of Birmingham, Academic Department, Birmingham Women’s Hospital, Mindelsohn Way, Birmingham, UK, B15 2TG; email: email@example.com.
- Tim Korevaar, MD, PhD, can be reached at Erasmus University Medical Center, Doctor Molewaterplein 40, 3015 GD, Rotterdam, the Netherlands; email: firstname.lastname@example.org.
- Caroline T. Nguyen, MD, can be reached at Keck Medical Center of the University of Southern California, Division of Endocrinology, Diabetes and Metabolism, 1540 Alcazar St., CHP 204B, Los Angeles, CA 90033; email: email@example.com.
- Alex Stagnaro-Green, MD, MHPE, MHA, can be reached at the University of Illinois College of Medicine at Rockford, 1601 Parkview Ave., Rockford, IL 61107; email: firstname.lastname@example.org.
Disclosures: Davies, Dhillon-Smith, Korevaar, Nguyen and Stagnaro-Green report no relevant financial disclosures.