Abstract
Background:
Resistance to thyroid hormone (RTH) is a rare thyroid disorder characterized by elevated free thyroid hormones with non-suppressed thyrotropin (TSH). Guidelines for the management of pregnancy in patients with RTH are not well defined. Chorionic villus biopsy is sometimes proposed to manage treatment based on the genotype of the fetus.
Patient findings:
A woman with RTH (c1243C>T, pR320C mutation in the thyroid hormone receptor β (THRB gene)) associated with Hashimoto's thyroiditis (HT) had three successful pregnancies. During the pregnancies, the mother was treated with levothyroxine (LT4). She never underwent chorionic villus sampling. The babies had normal birth weights. The first child harbored the THRB mutation.
Summary
Introduction
R
The association between RTH and autoimmune thyroid disease (AITD) has been previously described (2). In RTH, increased thyroid hormone secretion can, in part, offset the tissue-dependent resistance to thyroid hormone action. Unfortunately, however, this is impossible in cases of ablative therapy or in destructive thyroiditis. In these patients, it is necessary to administer levothyroxine (LT4) in supraphysiological doses (1).
RTH has equal prevalence in both sexes (1,2). Transmission is usually autosomal dominant (2,8). The condition's prevalence is ∼1/40,000 live births (1,2). Pregnancies with this syndrome are therefore rare. Recommendations for the management of pregnancies of RTH patients are not well defined. The rate of miscarriage is increased three- to fourfold (1), especially for unaffected fetuses, because high levels of maternal thyroid hormones can induce a thyrotoxic state with low birth weight (9). Some authors suggest prenatal diagnosis should determine appropriate management, but this type of procedure is not without risk (10).
Below, we describe a woman with RTH and Hashimoto's thyroiditis (HT) whom we followed during three pregnancies with good outcomes. The first fetus harbored the mutation, while the second and the third babies were unaffected. The proposita refused chorionic villus sampling. We discuss the management of pregnancy in patients with RTH associated with Hashimoto's disease and the indication for chorionic villus biopsy.
Patient
This 32-year-old female patient had a goiter since her teenage years, and had been followed in our hospital since she was 25 years old. When first evaluated at our institution, TSH was 3.01 mIU/L (normal 0.2–3.5 mIU/L; sandwich immunoassay on DxI, Beckman Coulter, Brea, CA, sensitivity 0.003 mIU/L), free thyroxine (fT4) was 28.4 pmol/L (normal 12–22 pmol/L, sensitivity 3.23 pmol/L), and free triiodothyronine (fT3) was 6.8 pmol/L (normal 3.1–6.8 pmol/L, sensitivity 1.32 pmol/L). Antithyroid peroxidase antibodies (anti-TPO) and anti-thyroglobulin antibodies (anti-TG) were positive. TSH receptor antibodies were negative. A thyroid ultrasound and a Tc99m thyroid scintigraphy showed a homogeneous goiter. Genetic analysis performed on DNA extracted from peripheral blood revealed a mutation c1243C>T in exon 9 of the THRB gene, converting arginine to cysteine at position 320 (R320C) (OMIM 190160) (11,12). She had no prior medical history, but there was a family history of goiter (her mother and her maternal aunt) and subclinical thyroid dysfunction (her maternal great-grandmother), whereas three brothers and sisters were in good health. The mother refused genetic testing. No mutation in the THRB gene was found in the siblings. Triiodothyroacetic-acid (TRIAC) was then introduced at a rate of 1 mg twice daily in order to reduce goiter and symptoms of anxiety and nervousness. Three months after initiation of the treatment, TSH was 2.94 mIU/L, fT4 20.6 pmol/L, and fT3 8.3 pmol/L; the patient was clinically euthyroid.
Six months later, silent thyroiditis was diagnosed. Hyperthyroidism was confirmed by a suppressed TSH of 0.03 mIU/L, an elevated fT4 of 54.2 pmol/L, and an elevated fT3 of 12.5 pmol/L. TRIAC was then stopped, and methimazole was started. Two months later, she was hypothyroid with a TSH of 17.3 mIU/L, a fT4 of 20.6 pmol/L, and a fT3 of 8.9 pmol/L. Methimazole was discontinued, and LT4 was initiated at 50 μg per day. With this treatment, her TSH was 3.25 mIU/L, fT4 31.0 pmol/L, and fT3 8.3 pmol/L, quite similar to levels found at her first visit. The patient was clinically euthyroid.
In June 2008, she was pregnant for the first time with no prior miscarriage. She took prenatal vitamins. Her LT4 dose was progressively increased to 100 μg per day, based on monthly thyroid function tests. Her TSH receptor antibodies were negative. She refused chorionic villus biopsy. Treatment was adjusted in order to maintain an fT4 range within 20% of the upper limit of normal, as recommended (1) and illustrated in Table 1A. This recommendation is especially relevant in the event that the fetus does not carry the THRB mutation. As we had no prenatal diagnosis, we opted for a cautious approach. The pregnancy was uneventful. At 39 weeks and 1 day, the patient gave birth by vaginal delivery to a girl weighing 3.48 kg with a length of 52 cm. The infant's heart rate was 150 bpm. The girl's TSH on day 2 was 18.28 mIU/L, and her fT4 was 61.9 pmol/L, suggesting RTH. Anti-TPOAb were positive, but anti-TGAb were negative. Table 2A illustrates the TSH and fT4 levels of the baby from day 2 until 3 years of age. As expected, given biological data and the normal birth weight, genetic analysis confirmed that the girl harbors the same mutation as her mother. She is now 4 years old and needs beta blockers to control her heart rate. After delivery, the mother returned to the initial LT4 dosage of 50 μg per day.
Normal values are those of the laboratory and not the normal reference range for pregnancy.
TSH, thyrotropin.
Normal values from Lem et al. (31) expressed in 0±2 SDS (standard deviation score).
ND, nondisposable.
Five months after delivery, the mother developed postpartum thyroiditis (PPT), characterized by a transient thyrotoxicosis followed by hypothyroidism. Her TSH was low at 0.27 mIU/L, and her peripheral hormones were elevated with a fT4 of 40.0 pmol/L and a fT3 of 8.7 pmol/L. LT4 was discontinued. One month later, she was hypothyroid. Her TSH was 27.42 mIU/L, fT4 was 10.3 pmol/L, and fT3 was 4.2 pmol/L. LT4 was reintroduced and progressively increased to 150 μg per day.
In March 2011, the mother became pregnant for the second time and started taking prenatal vitamins. She once again opted not to undergo chorionic villus biopsy. Therapy with LT4 was progressively increased to a level of 200 μg per day and decreased to 175 μg in week 34 of her pregnancy. Her fT4 level was maintained in the same range as during the first pregnancy, as illustrated in Table 1B. TSH receptor antibodies were negative. At 38 weeks and 5 days, she gave birth by vaginal delivery to a girl with a birth weight of 3.53 kg and a length of 52 cm. The baby's heart rate was 119 bpm. The girl's TSH at day 5 was 3.29 mIU/L, her fT4 was 23.2 pmol/L, and her FT3 was 4.1 pmol/L. Antibodies were not tested. The baby's weight was normal, and her TSH was not suppressed (Table 2B). Mutational analysis for the THRB gene mutation was negative, and a clinical evaluation at 2 months of age showed a euthyroid status. Postpartum, the LT4 substitution of the mother was reduced to 125 μg per day.
Six months after delivery, the mother again developed PPT. Her TSH was 0.04 mIU/L, fT4 was 42.6 pmol/L, and fT3 was 8.0 pmol/L. Therapy with LT4 was discontinued. As after the first delivery, she subsequently developed hypothyroidism 1 month later. Her TSH was 132 mIU/L, fT4 was 5.2 pmol/L, and fT3 was 3.8 pmol/L. LT4 was started again and increased to 150 μg per day. She was well controlled with this dose until the following pregnancy.
In June 2013, the proposita became pregnant for the third time. LT4 was increased up to 200 μg per day, and decreased to 175 μg in week 35. fT4 was maintained between 19.3 and 27.1 pmol/L, as illustrated in Table 1C. TSH receptor antibodies were still negative. The third delivery occurred in February 2014 at 40 weeks and 1 day. She gave birth by vaginal delivery to a girl weighing 4.050 kg and with a length of 54 cm. The newborn's heart rate was 130 bpm. The girl's TSH at birth was 4.34 mIU/L, her fT4 was 12.9 pmol/L, and fT3 was 2.6 pmol/L (cord blood determinations) as shown in Table 2C. Anti-TPOAb were positive. Screening for the mutation was negative. The mother's thyroid hormone substitution was reduced to 150 μg of LT4.
Discussion
We discuss the management of a woman with RTH and HT during three successful pregnancies. The woman never had a miscarriage. This case is particularly well documented, especially because the pregnancies occurred in close proximity. Genetic analyses were performed on the offspring, and we have a 4-year follow-up for the first child. To our knowledge, this is the first such case to be described in the literature. This patient features RTH associated with HT and consecutive episodes of thyroiditis with two episodes of PPT.
Several situations may be encountered in pregnant patients with RTH (1). In the first scenario, both the mother and the fetus exhibit RTH. This situation is typically characterized by a normal birth weight and normal newborn TSH in the offspring. No treatment is recommended during pregnancy, given that the fetus tolerates elevated circulating free thyroid hormones because it harbors the mutation and is resistant to thyroid hormones. In the second scenario, the mother exhibits RTH, whereas the fetus does not harbor the mutation. This scenario is typically characterized by a low birth weight and hyperthyroidism of the fetus because it is subjected to the potentially deleterious effects of high free thyroid hormone levels. Fetal loss is more frequent for mothers with RTH carrying an unaffected fetus (13). Given this scenario, the goal is to maintain a fT4 value within 20% of the upper limit of normal. This can be achieved using antithyroid drugs, but attention must be paid to avoid fetal hypothyroidism. In this case, the fetal genotype must be known to allow for proper pregnancy management. In the third scenario, the mother exhibits RTH but has undergone prior ablative treatment (radioiodine, surgery) or destructive thyroiditis (as illustrated by our patient), while the fetus harbors the mutation. In this situation, the outcome consists of a normal birth weight and euthyroid status in the fetus, as illustrated here.
The mutation found in our patient (R320C) has previously been described in six families (12,14 –18). It tends to generate a moderate phenotype compared to other mutations (19). Two unrelated families presenting this mutation have been compared, and genetic factors appear to modulate the expression of the RTH phenotype. The affected individuals of one family exhibited small goiters and an attention deficient hyperactivity disorder, while these events were absent in the subjects carrying the mutation in the other family (19). Autoimmune disease was also present in both families (19). The combination of RTH and AITD is more prevalent in RTH than in control groups (first-degree relatives unaffected by RTH) (2). It is well known that the presence of thyroid antibodies in the first trimester of pregnancy increases the risk of developing PPT by 50% (20). Hence, it is not surprising that our patient had an episode of PPT after each pregnancy, especially since the recurrence rate is so high (75%) (20). Paragliola et al. described a woman with RTH (due to a V283A THRB mutation) who had two pregnancies with favorable outcomes. She required no treatment but presented episodes of PPT after the first two pregnancies (21).
As RTH is a rare disease, successful pregnancies are worth reporting in order to improve management recommendations and thus avoid miscarriages. In the literature, there are only scarce case reports of pregnancies associated with RTH (with A317T, R243W, R243Q, T337A, V283A, M310L, T329N, or G339S mutations) (1,8,12,21 –25). Of the 14 reported pregnancies (1,8,21 –25,26 –28) with RTH, prenatal diagnosis was only performed in two cases (1,22). Weiss et al. described a woman affected by RTH (due to an A317T THRB mutation) who benefited from prenatal diagnosis confirming that the fetus did not harbor the disease. That patient was treated with propythiouracil (PTU) during the pregnancy in order to decrease fT4 to a level of 25% above the upper limit of normal. She gave birth to a baby of normal weight, and the infant's thyroid tests fell into the normal range. At 4 days of life, the girl had to be placed on LT4 due to increased TSH levels, probably due to the transplacental crossing of PTU (1). Asteria et al. described the case of a woman with RTH (due to a T337A THRB mutation) treated with TRIAC until early pregnancy. At 17 weeks, a chorionic villus biopsy confirmed that the fetus was affected by RTH. At 20 weeks of pregnancy, TRIAC was reinstated following symptoms of hyperthyroidism in the mother and because the fetus carried the mutation. At 29 weeks of pregnancy, the fetus had developed growth retardation and goiter. At 33 weeks, following a cordocentesis complication, the mother had to undergo an emergency cesarean section. At birth, the baby developed multiple-organ failure and respiratory distress syndrome, as well as transient hypothyroidism (22). To our knowledge, no pregnancy in a patient with the R320C mutation has been reported so far.
To optimize the management of the pregnancies in our RTH patient, we tried to detect subtle differences in the mother's and children's clinical or biological parameters. Unfortunately, no information on fetal growth and heart rate during the pregnancies are available. Monitoring of fetal growth and heart rate should be done during pregnancy to detect adverse effects of thyroid hormone excess on the fetus, particularly when the genotype is unknown. We found a slightly lower birth weight (3.48 kg vs. 3.53 kg and 4.05 kg) in the child carrying the mutation, as well as tachycardia, compared to the two unaffected siblings. The only observed difference in the mother concerned the prescribed LT4 doses during the second and the third pregnancy (150–200 μg), which were proportionally lower (25%) than in the first pregnancy (50%) (50–100 μg). The peculiarity of this case consists in the fact that the mother's RTH was associated with destructive thyroiditis resulting in hypothyroidism. Because autoimmune thyroiditis had partially destroyed the thyroid, the thyroid hormone reserve was lower than in a normal pregnancy. A 50% increase in the initial LT4 dosing is usually recommended in pregnancies of hypothyroid women based on the physiologic increase in maternal thyroid hormone production; this recommendation was applied during the first pregnancy, whereas the initial dose was increased by 25% for the second and the third pregnancy.
The more common challenge in RTH is how to manage high maternal thyroid hormone levels. Should thyroid hormones levels be adjusted depending on whether or not the fetus harbors the mutation? In this context, chorionic villus sampling followed by mutational analysis could be informative. Because we had no antenatal diagnosis, we tried to maintain fT4 levels within a maximum of 20% of the upper limit of the reference range during both pregnancies in order to avoid the potentially deleterious effects of overtreatment, which carries the consequence of low birth weight and inhibited TSH in normal fetuses. fT4 was maintained between 15.5 and 27.1 pmol/L during the first pregnancy and 20.6 and 27.1 pmol/L during the second and 19.3 and 27.1 pmol/L during the third one. This was made possible by monthly monitoring. In our case, fT4 would normally not exceed 24 pmol/L. For the first pregnancy, as the fetus harbored the mutation, she was not subjected to the deleterious effects of high thyroid hormone levels. For the second and the third pregnancy, the fetuses were normal and not exposed to high maternal fT4 levels, as TSH at birth was not inhibited. The association of RTH with ablative therapy or HT permits an easier management of the pregnancy than in RTH without ablative therapy or thyroiditis. Obviously, preventive thyroidectomy cannot be proposed to a woman with RTH planning pregnancy. The difficulty of the follow-up of a pregnant woman with RTH is the titration of the dose of LT4 to keep fT4 at maximum 20% above the upper limit of the reference range based on determinations of TSH and fT4, and acknowledging that there is no established reference range for TSH for pregnant women with RTH (29). TSH seems to be a better indicator of thyroid function in normal pregnant woman than fT4 (30).
Retrospectively, the successful outcomes of the three pregnancies reported here, including one affected and two unaffected fetuses, prompted us to question the usefulness of antenatal diagnosis. Would our therapeutic approach have changed based on prenatal genetic analysis? Some authors propose prenatal diagnosis using a chorionic villus biopsy in order to guide management (1,22,27). We offered prenatal diagnosis to avoid a possible overtreatment of unaffected fetuses. Our patient refused that procedure. The risk of miscarriage is especially high for unaffected fetuses subjected to the toxic effects of high thyroid hormone levels. One of the problems is that miscarriages often occur precociously, before a prenatal diagnosis has been made (by the 11th week). In addition, such genetic analyses are not performed in all hospitals, and results can be delayed. For example, our analysis is not performed in our clinic, but rather must be sent to another university center with a return of results 10 days later. Moreover, chorionic villus sampling is an invasive procedure that is not without risk and associated with a fetal loss rate between 0.4% and 1% (10).
Given the rarity of RTH, few cases have been studied during pregnancy to date. The lack of prospective studies makes it difficult to define management recommendations for pregnancy. The outcomes reported here suggest that prenatal diagnosis may not impact the therapeutic management of pregnant RTH woman who underwent prior ablative treatment or who have developed hypothyroidism. Overtreatment should be avoided, and fT4 levels should not be more than 20% above the reference range. Monitoring fetal heart rate can help the physician to suspect fetal hyperthyroidism.
Conclusion
In RTH patients, pregnancy should, if possible, be planned and must be closely followed. Our observations in this mother with RTH due to a R320C THRB mutation and a history of destructive thyroiditis suggest that prenatal diagnosis could be avoided. However, this specific situation should be adapted case by case depending on the severity of the mutation. We conclude that in patients with RTH associated with ablative therapy (including destructive thyroiditis), chorionic villus biopsy is not indicated. In this particular situation, the procedure's risks appear to outweigh the potential benefits. It is sufficient to maintain fT4 levels within a maximum of 20% above the upper limit of the reference range whether the fetus harbors the mutation or not. Pregnancy management might be easier in RTH cases associated with ablative therapy or destructive thyroiditis. Yet, prenatal diagnosis remains useful and appears warranted in RTH women without associated thyroiditis or with multiple miscarriages.
Footnotes
Author Disclosure Statement
The authors declare that no competing financial interests exist.
