Testing, Monitoring, and Treatment of Thyroid Dysfunction in Pregnancy

Abstract

Both hyperthyroidism and hypothyroidism can have adverse effects in pregnancy. The most common causes of thyrotoxicosis in pregnancy are gestational transient thyrotoxicosis and Graves’ disease. It is important to distinguish between these entities as treatment options differ. Women of reproductive age who are diagnosed with Graves’ disease should be counseled regarding the impact of treatment options on a potential pregnancy. Although the absolute risk is small, antithyroid medications can have teratogenic effects. Propylthiouracil appears to have less severe teratogenicity compared to methimazole and is therefore favored during the first trimester if a medication is needed. Women should be advised to delay pregnancy for at least 6 months following radioactive iodine to minimize potential adverse effects from radiation and ensure normal thyroid hormone levels prior to conception. As thyroid hormone is critical for normal fetal development, hypothyroidism is associated with adverse obstetric and child neurodevelopmental outcomes. Women with overt hypothyroidism should be treated with levothyroxine (LT4) to a thyrotropin (thyroid-stimulating hormone; TSH) goal of <2.5 mIU/L. There is mounting evidence for associations of maternal hypothyroxinemia and subclinical hypothyroidism with pregnancy loss, preterm labor, and lower scores on child cognitive assessment. Although there is minimal risk of LT4 treatment to keep TSH within the pregnancy-specific reference range, treatment of mild maternal thyroid hypofunction remains controversial, given the lack of clinical trials showing improved outcomes with LT4 treatment.

Case Presentation

A 27-year-old woman is referred to the endocrine clinic for evaluation of abnormal thyroid function tests. She has no significant past medical history and is not taking any medications. She recently saw her primary care provider for a 20-pound (9 kg) unintentional weight loss over 4 months, palpitations, anxiety, and irregular menstrual periods. Thyroid function tests done at that visit included serum thyrotropin (thyroid-stimulating hormone; TSH) <0.01 mIU/L (reference range, 0.35-4.9 mIU/L) and free thyroxine (T4) 2.4 ng/dL (reference range, 0.6-1.8 ng/dL). Her last menstrual period was 2 months ago. Her blood pressure is normal, and her heart rate is 100 bpm. Physical exam shows an anxious-appearing thin female, and is notable for tachycardia, a resting hand tremor, and a thyroid gland that is diffusely enlarged to about 45 grams without palpable nodules. There is no stare or evidence of exophthalmos. A spot urine pregnancy test comes back positive.

Interpretation of Thyroid Function Tests in Pregnant Women

Because of physiologic changes in thyroid hormone levels across gestation, nonpregnancy reference ranges should not be used in evaluating thyroid function tests in pregnancy. In early pregnancy, high serum levels of human chorionic gonadotropin (hCG) have weakly thyrotrophic effects on the thyroid gland, leading to increased thyroid hormone production and negative feedback to the pituitary (1). Thus, in the first trimester when serum hCG levels are highest, serum TSH tends to be lower compared with nonpregnant adults. A large study of 4800 Chinese pregnant women without any history of thyroid disease and negative thyroid antibody titers showed a downward shift of the serum TSH distribution at 7 to 12 weeks’ gestation compared with nonpregnant adults, with a reduction in the lower limit by 0.1 to 0.2 mIU/L and in the upper limit by 1.0 mIU/L of serum TSH relative to the nonpregnant TSH reference range (2). The American Thyroid Association (ATA) recommends using assay-specific, trimester-specific reference ranges derived in women without known thyroid disease, with optimal iodine nutrition, and with negative thyroid peroxidase (TPO) antibody. Currently, such ranges are not widely available. In the absence of such pregnancy-specific reference ranges, the ATA recommends using a lower limit of TSH of 0.1 mIU/L, about 0.4 mIU/L lower than the lower limit for nonpregnant adults, and an upper limit of TSH of 4 mIU/L, about 0.5 mIU/L lower than the upper limit for nonpregnant adults (3).

High estrogen levels in pregnancy increase serum thyroxine binding globulin (TBG) levels. Since the majority of thyroid hormone in circulation is bound to TBG, total triiodothyronine (T3) and T4 levels frequently exceed nonpregnancy upper normal limits in the absence of hyperthyroidism. TBG levels rise in a linear fashion from 7 to 16 weeks’ gestation. The upper limit of total T3 or T4 levels can be calculated by adding 5% per week to the nonpregnancy upper normal limits between 7 and 16 weeks’ gestation, after which TBG levels stabilize and the gestation-specific upper normal limits stabilize at 150% of the nonpregnancy upper limits (4). Indirect immunoassay methods for measuring free thyroid hormone levels are confounded by high levels of TBG. The free T4 index is less affected by high serum TBG and is considered more accurate than direct free T4 assays when TBG is elevated (5). Using either trimester-specific, assay-specific ranges for free T4, or the free T4 index, or total T4 is recommended for assessment of thyroid hormone levels in pregnancy (5).

Etiology of Thyrotoxicosis in Pregnancy

Thyrotoxicosis in pregnancy can occur due to hyperthyroidism (increased production of thyroid hormone) or increased release of preformed hormones. The most common cause of hyperthyroidism in young women is Graves’ disease, with a prevalence that ranges from 0.4% to 1.0% before pregnancy and is 0.2% during pregnancy (6). The most common cause of nonautoimmune hyperthyroidism in the first trimester is gestational transient thyrotoxicosis (GTT), which occurs in 2% to 11% of pregnancies, with a higher prevalence in Asian women (7, 8). It is important to distinguish GTT from Graves’ disease, since treatments for the 2 entities are different. Painless thyroiditis and subacute thyroiditis, in which thyrotoxicosis results from the increased release of preformed thyroid hormone from the inflamed thyroid gland, can be seen in young women of child-bearing age, but this is uncommon during pregnancy. Toxic adenoma is also infrequent in pregnant women.

Gestational Transient Thyrotoxicosis

GTT presents with suppressed TSH and elevated T4 levels in early pregnancy. It is mediated by high serum hCG levels. Because hCG has a weak affinity for thyrotropin receptor on thyroid follicular cells (9), elevated hCG stimulates thyroid hormone production. Thus, very high levels of hCG, or particularly bioactive hCG can lead to GTT. Because high hCG levels are also frequently associated with nausea, vomiting, and hyperemesis gravidarum (10), women with GTT usually describe significant nausea and emesis. The diagnosis of GTT is made clinically. Pregnant women with GTT typically do not recall having symptoms of thyrotoxicosis preconception and will not have stigmata of Graves’ disease such as goiter or exophthalmos. Serum hCG levels are not useful in differentiating between GTT and Graves’ disease (11). The ratio of T3 to T4 is often higher in Graves’ disease than in GTT, and TSH receptor antibody (TRAb) levels are typically elevated only in Graves’ disease. As the name implies, GTT is a transient condition, since serum hCG levels peak at around 8 to 10 weeks’ gestation and then decline. GTT is not associated with adverse pregnancy outcomes (12, 13), so treatment with antithyroid drugs (ATDs) is not indicated. Patients with GTT should be provided with supportive care for severe nausea and vomiting, and any electrolyte imbalance or volume depletion. Pregnancy-safe beta-blockers such as labetalol, using the lowest effective dose starting at 100 mg twice a day, can be considered if thyrotoxic symptoms are severe. It is worth noting that atenolol should be avoided as it is associated with fetal growth retardation (14).

Graves’ Disease

During gestation, typical symptoms of hyperthyroidism such as palpitations, fatigue, or heat intolerance, may be confused with symptoms of normal pregnancy. Pregnant women with Graves’ disease may have a diffusely enlarged goiter or Graves’ ophthalmopathy. They may report an inability to gain weight despite good appetite. Suggestive signs or symptoms should prompt screening for hyperthyroidism. In overt hyperthyroidism from Graves’ disease, serum TSH is suppressed and T3 and/or T4 levels are elevated above gestation-specific reference ranges, with a preferential production of T3 relative to T4. TRAb levels are almost invariably elevated. Unregulated stimulation of thyroid cells by TRAb is the hallmark of Graves’ disease, although TRAb can be either stimulating or blocking to the TSH receptor. In measuring TRAb, either a competitive binding immunoassay or a functional cell-based bioassay can be used. TSH receptor–binding inhibitory immunoglobulin (TBII), a binding immunoassay, detects TRAb in serum by testing its ability to bind to TSH or TSH receptor monoclonal antibodies (15). Although highly sensitive (97.4%) and specific (99.2%) for Graves’ disease (16), the binding assay does not distinguish between blocking and stimulating antibodies. Thyroid-stimulating immunoglobulin (TSI) is a stimulating TRAb measured using a bioassay that measures the levels of cyclic adenosine monophosphate (cAMP) induced by TRAb binding to the TSH receptor (15). Because the bioassay measures the functional activity of the TRAb, the level correlates with disease activity in Graves’ hyperthyroidism. In addition to providing an accurate diagnosis of the etiology of thyrotoxicosis, TSI levels can be used to determine the risk of thyroid dysfunction or goiter development in the fetus or newborn, as discussed below. If a TSI assay is not available, a TRAb assay can be used together with the clinical picture and thyroid hormone levels to make the diagnosis of Graves’ disease and determine fetal goiter risk.

Further evaluation of the patient

Laboratory testing obtained after initial evaluation in the endocrine clinic included: TSH <0.01 mIU/L (nonpregnancy reference range, 0.35-4.9 mIU/L), total T4 37.7 mcg/mL (nonpregnancy reference range, 5.1-11.4 mcg/mL), T3 uptake 18.4% (nonpregnancy reference range, 21%-38%), FT4 index 9.8 (nonpregnancy reference range, 1.0-4.0), and total T3 >400 ng/dL (nonpregnancy reference range, 8.3-160 ng/dL). The TSI level was 300 (reference range, ≤140%). The patient was diagnosed with overt hyperthyroidism due to Graves’ disease and was started on propylthiouracil (PTU) 50 mg 3 times daily.

Treatment options for Graves’ disease in pregnancy

The degree of hyperthyroidism from Graves’ disease in pregnancy may be mild or severe. Subclinical hyperthyroidism (low serum TSH with normal-for-gestation thyroid hormone levels) does not appear to have adverse effects in pregnancy (17), and monitoring without treatment is recommended. However, moderate-to-severe hyperthyroidism should be treated in order to avoid the risks of preeclampsia, pregnancy loss, maternal and fetal congestive heart failure, thyroid storm, premature delivery, intrauterine growth retardation, and low birth weight (18-21).

ATD treatment is required when Graves’ disease causes overt hyperthyroidism in pregnancy. The ATDs methimazole (MMI) and PTU are effective in treating hyperthyroidism, but both may be associated with adverse effects. Both ATDs can increase serum liver enzyme levels, which generally normalize after discontinuation of medications (22, 23). Hepatotoxicity from MMI has been thought to be mediated via cholestasis (24-26), while hepatotoxicity from PTU was thought to be from hepatocellular damage. More recent studies have shown no significant difference in the prevalence of cholestasis between the 2 ATDs, although PTU has been associated with a higher incidence of more severe liver injury (22, 23). Although very rare (prevalence 1:10 000), PTU can lead to fulminant hepatic failure, including in pregnant women (26, 27). Outside the setting of pregnancy, MMI is generally preferred over PTU for this reason. During the period of organogenesis in early pregnancy, the teratogenicity of ATDs must be considered because both MMI and PTU cross the placenta (28). MMI has been known to cause birth defects since the 1970s (29), while for decades PTU was thought to be safe in early pregnancy (30). Recent studies have shown that both MMI and PTU increase risks of congenital malformations, although the absolute risk is small (31). In a Danish nationwide registry-based cohort study of approximately 800 000 mother-child pairs, 8% of children exposed to PTU, 9.1% children exposed to MMI, and 10.1% of children exposed to both MMI and PTU in utero had birth defects, compared with about 5% of children who were not exposed to ATDs or who were born to mothers without thyroid dysfunction (31). This increase in risk was also seen in a larger nationwide registry-based cohort study of Danish women; the risks of birth defects were 9.6% in those exposed to MMI in utero and 8.3% in those exposed to PTU in utero, compared with 6.7% in those nonexposed (32). Similarly, increased risk of congenital malformations was found in a large Korean study (33). Exposure to MMI in utero is associated with more severe birth defects, such as esophageal or choanal atresia, abdominal wall defects, aplasia cutis, and ventricular septal defects (31, 34). Exposure to PTU in utero has been associated with less severe birth defects, such as cystic malformations of the face or neck, or hydronephrosis or cysts in the urinary tract (31, 35). Consequently, PTU is recommended for treatment of Graves’ disease during the period of organogenesis in the first trimester. Whether or not to change from PTU to MMI in the second trimester is unclear. In doing so, the PTU-associated risk of fulminant hepatic failure must be weighed against the risk of causing or exacerbating thyroid dysfunction with the change. If a decision is made to change PTU to MMI, an approximate dose ratio of 20:1 can be used (eg, PTU 100 mg twice a day would be equivalent to MMI 10 mg a day) (3).

Other treatment modalities for Graves’ disease are less favored in pregnancy. Radioactive iodine (RAI) ablation is absolutely contraindicated in pregnancy (36). In severe cases of Graves’ disease that are inadequately controlled with medical treatment or in patients unable to tolerate ATDs, thyroidectomy can be considered. If surgery is needed, the second trimester is safest, as it avoids potential teratogenic effects of anesthesia on the fetus in the first trimester and risk of inducing premature labor in the third trimester (3).

Monitoring thyroid function in pregnant women with Graves’ disease

It is important to avoid overtreatment of thyrotoxicosis in pregnancy. During the critical period of development in early gestation, the fetus depends on maternal T4 crossing the placenta for its thyroid hormone requirement. ATDs cross the placenta and therefore can affect the fetal thyroid gland once it is functional around mid-gestation (37). The fetal thyroid gland is more sensitive to the effects of the ATDs than the maternal thyroid gland, as demonstrated by a Japanese study which showed that fetal thyroid function may be low even when ATD-treated mothers are euthyroid (38). Therefore, it is recommended to employ the minimal ATD dose that will keep maternal free T4 at or just above the upper limit of the normal range and TSH no higher than the lower end of the reference range. If using total T4 for monitoring, a goal total T4 level around 1.5 times of the upper limit of the nonpregnancy reference range can be used. A maternal serum TSH level at the upper pregnancy reference range is an indication for reduction in the ATD dose to prevent overtreatment. Since ATDs cross the placenta to a much greater extent than levothyroxine (LT4), the “block-and-replace” strategy using both ATD and LT4 is not recommended in pregnancy (3). Overtreatment of maternal Graves’ disease can lead to fetal hypothyroidism with a reciprocal increase in fetal TSH, increasing risk of fetal goiter development.

Maternal TRAb crosses the placenta and can cause fetal hyperthyroidism (manifested as fetal tachycardia, intrauterine growth retardation, and an increase in bone mineralization) once the fetal thyroid is functional. TRAb may also cause fetal goiter, potentially leading to respiratory difficulty after birth (39). Although TRAb titers decline after thyroidectomy, they may remain elevated for years. Following radioactive iodine ablation, TRAb titers rise above baseline for about 1 year and may remain detectable for years. Thus, it is recommended to check the TSI levels during early pregnancy in pregnant women with history of active treated Graves’ disease or prior thyroidectomy or radioactive iodine ablation. If the TSI level is greater than 2.5 to 3 the times upper limit of the reference range, it should be rechecked at 18 to 22 weeks’ gestation (40, 41). If TRAb or TSI levels remain elevated in the second trimester, patients should be managed in close collaboration with maternal-fetal medicine and neonatology. Fetal ultrasound can be used to monitor for goiter (which can be associated with both fetal hyper- and hypothyroidism) and signs of fetal hyperthyroidism (42, 43).

Patient’s disease course

Unfortunately, the patient experienced a spontaneous pregnancy loss at about 9 weeks’ gestation. She returns for follow-up after her pregnancy loss and asks about options for treatment of Graves’ disease prior to conceiving again. Her hyperthyroidism has worsened postpregnancy, necessitating an increase in her PTU dose. She would like to avoid taking ATDs during future pregnancies.

Preconception counseling in women with Graves’ disease

Given the risk of obstetric complications from uncontrolled hyperthyroidism and the potential adverse effects of ATDs in pregnancy, potential pregnancy should be discussed with all women of reproductive age who have Graves’ disease. Hyperthyroid women should be counseled to postpone pregnancy until euthyroidism is achieved. For women who prefer medical therapy, PTU may be preferable once a woman is actively trying to conceive, in order to avoid any exposure to MMI in early gestation. In a subset of women requiring low-dose ATDs (less than 5-10 mg of MMI/day or less than 100-200 mg of PTU/day) to maintain euthyroidism, those who have been treated for at least 6 months, those who do not have active Graves’ ophthalmopathy, or those with low or negative TRAb titers, ATDs may be stopped as soon as pregnancy is confirmed in the hope that hyperthyroidism will not recur during pregnancy or will not recur until after the first trimester, when ATD use is safer (3). Thyroid function should be carefully monitored in these women (initially every 2 weeks) after stopping ATDs to monitor for possible recurrence of overt hyperthyroidism.

For women who prefer to avoid ATDs during pregnancy altogether, preconception definitive therapy with either RAI ablation or thyroidectomy can be considered. Potential complications of RAI ablation, including worsening of Graves’ ophthalmopathy, should be discussed with the patient. TRAb levels may transiently increase from baseline and remain elevated following RAI ablation for many years, while they decrease gradually following thyroidectomy, generally within the first year (44). Pregnancy should be delayed at least 6 months after RAI ablation to minimize potential adverse effects from radiation (45) and to ensure that women are euthyroid and on a stable LT4 dose (36).

Patient course, continued

The patient opted for treatment with RAI ablation and received 15 mCi of iodine 131. She was advised not to become pregnant for 6 to 12 months. PTU was discontinued 2 months after her RAI ablation and LT4 was initiated. She returns for follow-up 7 months after RAI ablation and reports that she just had a positive home pregnancy test. Her last menstrual period was 6 weeks ago. One month ago, her serum TSH was 1.8 mIU/L on LT4 75 mcg daily.

Hypothyroidism in Pregnancy

Effects of Overt Hypothyroidism in Pregnancy

Overt maternal hypothyroidism in pregnancy is associated with multiple adverse obstetric outcomes. Davis et al reported increased risks of preeclampsia and placental abruption leading to low birth weight and fetal death in overtly hypothyroid pregnant women (18). Leung et al showed an increased risk of gestational hypertension in hypothyroid pregnant women, leading to increased risks of preterm delivery and low birth weight in babies (46). Allan et al reported a 4.4-fold increase in risk of fetal death in women with TSH ≥6.0 mIU/L relative to euthyroid mothers (3.8% vs 0.9%) (47). Increased risks of fetal loss and premature delivery in pregnant women with inadequately treated hypothyroidism (48), and increased risk of preeclampsia and resultant increase in risk of premature delivery (19) have also been reported.

Thyroid hormone affects neuronal migration, connection, myelination, and synaptogenesis (49, 50). Thyroid hormone receptors are also expressed during cortex development (51, 52). These findings support an important role for thyroid hormone in normal brain development. Because the fetal thyroid gland does not mature until 18 to 20 weeks’ gestation (37), thyroid hormone necessary for early development comes entirely from transplacental passage of maternal hormone. Haddow et al conducted one of the first studies showing significant effect of maternal hypothyroidism on subsequent neurodevelopment of children. In this case-control study, children of hypothyroid mothers had 4-point lower IQ at 7 to 9 years of age compared with children of euthyroid mothers (53). When hypothyroid women who were treated with LT4 in pregnancy were excluded, this difference in child IQ increased to 7 points (53).

Effects of Subclinical Hypothyroidism and Hypothyroxinemia in Pregnancy

Subclinical hypothyroidism is defined as an elevated serum TSH in the setting of normal free T4. Isolated maternal hypothyroxinemia is defined as normal serum TSH level with a decrease in serum free T4 level to below the 2.5th to 5th percentile of the reference range. Whether or not to screen for or treat subclinical hypothyroidism and hypothyroxinemia in pregnancy has been highly controversial.

Studies examining the associations between maternal subclinical hypothyroidism and adverse pregnancy outcomes have had mixed results. In a prospective cohort study of almost 11 000 women in the United States, no significant associations were found between maternal TSH elevation above the 97.5th percentile in the first or second trimesters and adverse pregnancy outcomes including pregnancy loss, gestational diabetes or hypertension, preeclampsia, preterm labor, and low or high birth weight (54). A 2015 meta-analysis of 10 cohort studies showed no significant association between maternal subclinical hypothyroidism and preterm birth (55). On the other hand, in a prospective cohort study of 3147 Chinese women, those with serum TSH values between 5.22 and 10 mIU/L had a risk of pregnancy loss which was 3.4 times greater compared to women with TSH <2.5 mIU/L (56). The risk of pregnancy loss was even higher when women also had thyroid autoimmunity: 5 times higher in women with positive thyroid antibodies and TSH levels of 2.5 to 5.22 mIU/L and 9.6 times higher in women with positive thyroid antibodies and TSH levels of 5.22 to 10 mIU/L (56). A meta-analysis of 18 cohort studies using trimester-specific TSH cutoffs from the 2011 ATA guidelines (2.5 mIU/L in the first trimester and 3 mIU/L in the second and third trimesters (57)) reported a significant association between maternal subclinical hypothyroidism and pregnancy loss (relative risk, 2.01; 95% CI, 1.66-2.44) (58). Most recently, a meta-analysis of individual data from about 47 000 women from 19 cohorts showed an increased risk of preterm birth in women with either subclinical hypothyroidism or isolated hypothyroxinemia (odds ratio [OR] 1.29; 95% CI, 1.01-1.64 and OR 1.46; 95% CI, 1.12-1.90, respectively) (59). These conflicting results may be partly due to variability across studies in the timing of TSH measurements and inclusion of TPO antibody assessment, but mounting evidence suggests that there are associations between maternal thyroid hypofunction and pregnancy loss and preterm birth.

Maternal subclinical hypothyroidism and hypothyroxinemia also appear to be associated with adverse child neurodevelopmental outcomes. A prospective cohort study of 3839 Dutch mother-child pairs showed an inverted U-shaped association between maternal FT4 levels and children’s IQ assessed at 6 years of age, suggesting that both low and high thyroid hormone exposure in utero can adversely affect child neurodevelopment (60). A similar association was not seen with maternal TSH levels. A recent meta-analysis of 11 cohort studies showed an increased risk of intellectual impairment in children born to mothers with either subclinical hypothyroidism or hypothyroxinemia (OR 2.14; 95% CI, 1.20-3.83 and OR 1.63; 95% CI, 1.03-2.56, respectively) (61).

Only a few clinical trials have assessed the effects of LT4 treatment in subclinically hypothyroid or hypothyroxinemic pregnant women. Two small randomized clinical trials have shown a decrease in preterm delivery in LT4-treated pregnant women with subclinical hypothyroidism with TSH ≥4 mIU/L with or without TPO antibody positivity (62, 63). A recent meta-analysis of 6 clinical trials and 7 cohort studies showed a decreased risk of pregnancy loss (OR 0.78; 95% CI, 0.66-0.94) and increased risk of preterm labor (OR 1.82; 95% CI, 1.14-2.91) in women with subclinical hypothyroidism treated with LT4 (64). The heterogeneity of results may be due to differences in LT4 doses used, definitions of subclinical hypothyroidism in each study, and timing of LT4 treatment initiation.

The effects of LT4 treatment for maternal subclinical hypothyroidism and hypothyroxinemia on child neurodevelopmental outcomes are unclear. In the Controlled Antenatal Thyroid Study (CATS), 390 women with subclinical hypothyroidism or isolated hypothyroxinemia were randomized to treatment with LT4 150 mcg/day. At 3 years of age, IQ of their children did not differ from that of children of 404 women with untreated subclinical hypothyroidism or hypothyroxinemia (65). These results persisted when children were tested at 9 years of age (66). Although this study showed no benefit of LT4 treatment of maternal subclinical hypothyroidism or hypothyroxinemia on childhood cognitive function, the findings are limited by the high dose of LT4 used (resulting in possible overtreatment) and initiation of treatment at a median of 12.3 weeks’ gestation, after the critical period of early development. Another randomized clinical trial of LT4 treatment of US pregnant women with subclinical hypothyroidism or isolated hypothyroxinemia showed similar results (67). In this study, 677 pregnant women with subclinical hypothyroidism and 526 pregnant women with isolated hypothyroxinemia were randomized to either LT4 treatment (100 mcg/day for subclinical hypothyroidism and 50 mcg/day for hypothyroxinemia) or placebo at a mean of 17.8 weeks’ gestation. Thyroid function was monitored throughout pregnancy to maintain TSH between 0.1 and 2.5 mIU/L. There were no significant differences in mean IQ scores of children measured at 5 years of age, in either the subclinical hypothyroidism or hypothyroxinemia group (67). There were also no significant differences in adverse obstetric outcomes in LT4-treated women compared those who were not treated, including preterm birth, gestational hypertension, preeclampsia, gestational diabetes, and pregnancy loss (67). An important limitation of this study is that LT4 treatment was initiated at median of 16.7 weeks’ gestation, well into the second trimester and possibly beyond the critical period for neurodevelopment.

Treatment of hypothyroidism in pregnancy

Pregnant women with overt hypothyroidism should be treated with LT4 to prevent obstetric complications and adverse child neurodevelopmental outcomes. LT4 monotherapy is recommended for treatment of hypothyroidism in pregnancy. Since T3 does not cross the placenta, use of T3, T3/T4 combination therapy, or desiccated thyroid may lead to inadequate thyroid hormone availability to the fetus despite normal maternal thyroid function (3).

LT4 treatment for subclinically hypothyroid or hypothyroxinemic pregnant women remains controversial. The ATA recommends evaluating pregnant women with TSH >2.5 mIU/L for the presence of TPO antibody and has tiered recommendations for the treatment of maternal subclinical hypothyroidism based on the degree of TSH elevation and the presence of thyroid autoimmunity (3). On the other hand, a recent American College of Obstetrics and Gynecology (ACOG) guideline provides a more conservative approach (68). While the ATA guideline favors treatment of maternal subclinical hypothyroidism for TSH values >4 mIU/L if the TPO antibody is positive, and for TSH >10 mIU/L if the TPO antibody negative, the ACOG guideline recommends treatment only if the free T4 level is low, essentially advocating treatment only for overt hypothyroidism. Specific recommendations from the 2 organizations are shown in Table 1. Treatment for isolated maternal hypothyroxinemia is not recommended, given the lack of treatment benefit in clinical trials (3).

Table 1.

Recommendations From ATA and ACOG Regarding the Treatment of Maternal Hypothyroidism and Hypothyroxinemia in Pregnancy(3, 68)

Laboratory data	ATA	ACOG
TPO antibody negative and TSH >10 mIU/L	Treat with LT4	Treat with LT4 only if free T4 is low
TPO antibody positive and TSH > pregnancy-specific range (or 4 mIU/L)	Treat with LT4	Treat with LT4 only if free T4 is low
TPO antibody positive and 2.5 mIU/L < TSH < pregnancy-specific reference range (or 4 mIU/L)	Consider treatment with LT4	No treatment
TPO antibody negative and pregnancy-specific reference range (or 4 mIU/L) < TSH < 10 mIU/L	Consider treatment with LT4	Treat with LT4 only if free T4 is low
Isolated maternal hypothyroxinemia	No treatment	Not discussed

Abbreviations: ACOG, American College of Obstetrics and Gynecology; ATA, American Thyroid Association; LT4, levothyroxine; TPO, thyroid peroxidase.

When LT4 treatment is warranted, the goal serum TSH is <2.5 mIU/L (3, 68). Most women who are on LT4 pre-pregnancy will need a 20% to 30% increase in their doses in pregnancy in order to maintain euthyroidism, largely due to the pregnancy-induced rise in TBG (69). Increase in TBG leads to an increased binding of thyroid hormone to TBG, which in turn decreases free thyroid hormone levels. This causes an increase in serum TSH levels via negative feedback to the pituitary gland. Women with hypothyroidism may not be able to increase thyroid hormone production appropriately in response to the rise in TSH, and therefore need a higher dose of LT4 in pregnancy. In LT4-treated women, serum TSH levels should be monitored every 4 weeks until mid-gestation (70) and then at least once around 30 weeks’ gestation to ensure normal thyroid hormone levels throughout pregnancy. Women who require an increase in their LT4 doses during pregnancy can decrease back to their pre-pregnancy dose immediately after delivery with a plan for repeat TSH testing at approximately 6 weeks’ postpartum.

Patient case conclusion

The patient was advised to empirically increase her LT4 dose to 100 mcg/day with a plan for TSH testing every 4 weeks through week 16 of gestation.

Take-home points:

• Gestational transient thyrotoxicosis and Graves’ disease are the most common causes of thyrotoxicosis in pregnancy.

• Although the absolute risk is small, antithyroid drugs can cause congenital malformations. If needed, propylthiouracil is preferred to methimazole in the first trimester.

• Overtreatment of thyrotoxicosis should be avoided in pregnancy, as it may lead to fetal hypothyroidism and goiter.

• Women should wait at least 6 months before conception after radioactive iodine ablation.

• Thyroid receptor antibody levels should be checked in early pregnancy in women with a history of active or treated Graves’ disease.

• Adequate thyroid hormone is critical for normal fetal development and there is an increase in thyroid hormone requirements in pregnancy.

• Women with hypothyroidism should be treated with levothyroxine for a goal TSH <2.5 mIU/L in pregnancy.

• Treatment of subclinical maternal hypothyroidism remains controversial. However, there is increasing evidence for associations between maternal thyroid hypofunction and adverse pregnancy and childhood neurodevelopmental outcomes.

Glossary

Abbreviations

ACOG

American College of Obstetrics and Gynecology

ATA

American Thyroid Association

ATD

antithyroid drugs

GTT

gestational transient thyrotoxicosis

hCG

human chorionic gonadotropin

LT4

levothyroxine

MMI

methimazole

OR

odds ratio

PTU

propylthiouracil

RAI

radioactive iodine

T3

triiodothyronine

T4

thyroxine

TBG

thyroxine binding globulin

TPO

thyroid peroxidase

TRAb

TSH receptor antibody

TSH

thyrotropin (thyroid-stimulating hormone)

TSI

thyroid-stimulating immunoglobulin

Additional Information

Disclosures: The authors have nothing to disclose.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.