An important clinical concept is that the risk of complications for mother and child is directly related to the duration and adequate control of maternal hyperthyroidism (see Table 14-13). The most common complication is gestational hypertension, with a risk of preeclampsia that is ~5-fold greater in women with uncontrolled hyperthyroidism. Other obstetrical complications include miscarriage, fetal malformation, placenta abruptio, preterm delivery and low birth weight [279, 286-290]. With regard to the risk of miscarriage in women with elevated thyroid hormone levels, a recent study showed that women with a genetic resistance to thyroid hormone (hence, who were euthyroid but had elevated T4 levels) experienced a significantly increased miscarriage rate compared to unaffected couples [291]. The hypothesis was that the elevated maternal thyroid hormone levels caused hyperthyroidism in the fetus not carrying the gene for thyroid hormone resistance. Congestive heart failure may occur in women left untreated or treated only for a short period of time in the presence of gestational hypertension or operative delivery [285, 292]. In a recent study from Australia, the outcome of pregnancy was evaluated in women with unrecognized maternal Graves' disease [293]. Results showed severe prematurity (mean gestational age of 30 weeks at delivery) associated with very low birth weight (<2 Kg) and neonatal hyperthyroidism requiring treatment with ATD. In contrast, for those patients in whom the diagnosis was made early and the treatment started promptly, the outcome was excellent. In a report of a family presenting non-autoimmune hyperthyroidism due to an activating TSH receptor gene mutation, it was recently shown that premature delivery and low birth weight were consistent features associated with this unusual cause of thyrotoxicosis [294]. Finally, in another study of 230 pregnant women with Graves’ disease in Japan, no adverse impact on the outcome of pregnancy was found in patients with adequately treated Graves' disease [295].

Graves’ disease is an autoimmune disease and it is therefore important to consider that the production of TSH-receptor antibodies (TRAb) may fluctuate during pregnancy and, in turn, influence the clinical course of the disease. Since Graves’ disease tends to undergo immunologic remission during gestation, TRAb detection may depend upon gestational age at measurement [287]. The classical course of Graves’ disease during pregnancy frequently encompasses exacerbation of hyperthyroidism during the first trimester and a gradual improvement in the second half of gestation. Maternal production of TRAb may remain elevated, even after a prior thyroidectomy or thyroid ablation using radioiodine, or the apparent cure of the disease by antithyroid drug (ATD) therapy given several years before pregnancy. In euthyroid pregnant women who have previously received ATD for Graves’ disease but who are currently not receiving ATD treatment, the risk of fetal/neonatal thyrotoxicosis is negligible and, therefore, systematic measurement of TRAb is not mandatory. For a euthyroid pregnant woman (with or without thyroid hormone replacement therapy) who has previously been treated with radioiodine or undergone thyroid surgery for Graves’ disease, the risk of fetal/neonatal thyrotoxicosis depends upon the level of TRAb produced by the mother. As a result, TRAb should be measured in early pregnancy to evaluate this risk. For pregnant women who take ATD for the treatment of active Graves’ disease (assuming that ATD was started before or in early pregnancy), TRAb should be measured in the first and last trimesters. If TRAb titers have not substantially decreased during the second trimester, the possibility of fetal thyrotoxicosis should be considered [46, 59, 208, 296]. Radetti et al. showed that TRAb’s maternal transfer to the fetus in a hyperthyroid pregnant woman with active Graves’ disease caused a marked increase in the fetal/maternal TRAb ratio [297]. Thus, the fetal-placental unit gradually concentrates proportionally larger fractions of thyroid stimulating antibodies from maternal blood, particularly during the second half of gestation, hence underlining the importance to systematically search for fetal thyroid dysfunction in all clinical conditions where the mother may produce large amounts of TRAb. In a recent study by Peleg et al., it was confirmed that neonatal thyrotoxicosis occurred only in infants of mothers with high TRAb titers [298].

As indicated above, hyperthyroidism due to Graves' disease usually tends to improve gradually during gestation, although exacerbations can be observed in the first weeks. Several reasons may help explain this spontaneous improvement: a) the partial immunosuppressive state characteristic of pregnancy, with a progressive decrease in TRAb production [299]; b) the marked rise in maternal serum TBG levels that tends to reduce serum free T4 & T3 fractions; c) the obligatory iodine loss specific of pregnancy that may, paradoxically, constitute an advantage for women with Graves' disease; d) changes in cytokine production between normal and Graves’ disease pregnant women may also help explain the transient remission of the disease [300]; and e) finally, a suggestion was made that the balance between TRAb’s blocking and stimulating activities may be modified in favor of blocking antibodies [301, 302]. The reality of the latter finding has, however, been recently denied by another study from Japan [303].

Since TRAb may contain a mixture of different action types on the TSH receptor, there is room for confusion. All antibodies that can compete with TSH for binding to the TSH receptor are identified as thyrotropin-binding inhibitory immunoglobulins (TBII). These antibodies are measured using commercially available kits that record the percentage of inhibition of TSH binding to a membrane preparation of TSH receptors. The assay does not measure the ability of the antibody to stimulate the receptor, but binding and stimulation frequently go in parallel. Therefore this assay is often used as a surrogate for an assay of TSH receptor stimulating antibodies. Antibodies that stimulate the receptor can be measured by their ability to stimulate cAMP production in a preparation of cell membranes containing the TSH receptor. This assay is specific for the pathogenic antibody in Graves’ disease, but is not generally available in hospital or commercial laboratories. Some antibodies bind to the TSH receptor and inhibit the stimulating activity of TSH. Such antibodies may be clinically significant since they can cause hypothyroidism (including in the fetus), but they are usually measured only in a research setting. Their assay depends upon their ability to reduce cAMP production induced by TSH, in the setting of the stimulating assay described above [304].

Propylthiouracil (PTU), methimazole (MMI) and carbimazole (CMI) have been used during gestation. These medications inhibit thyroid hormone synthesis via inhibition of iodine organification and iodotyrosine coupling. Pregnancy itself does not appear to alter significantly the pharmacokinetics of ATD and both PTU and MMI (or CMI) appear equally effective [307]. The ATD dosage should be maintained at a minimum and drugs discontinued, when feasible (often possible after six months of gestation). Combined administration to the mother of ATD and thyroxine should be avoided, since the trans-placental passage of ATD is high, while it is negligible for thyroid hormones, and hence will not protect the fetus from ATD-induced hypothyroidism.

All ATD cross the placenta and may thus inhibit fetal thyroid function. PTU is more water soluble and more extensively bound to albumin at physiologic pH than MMI. Theoretically therefore, treatment with MMI may result in an increased trans-placental passage relative to PTU. These facts have led authors to recommend preferentially the use of PTU both during pregnancy and lactation, based on the concept that the fetus/infant might be at higher risk of developing hypothyroidism when women receive MMI. From all the available information taken together, however, differential placental transfer of PTU and MMI appears unlikely and by itself does not support the preferential use of PTU versus MMI. Furthermore, MMI (or CMI) is commonly used in pregnancy in countries where PTU is not commercially available (for instance, in Japan and several European countries), without particular problem. Therefore, the recommendation for the sole use of PTU does not appear to be justified. [ 75, 308-314].

General guidelines for the treatment of Graves’ disease in pregnancy are summarized in Table 14-14. The principle of therapy is to administer the lowest ATD dose needed for controlling clinical symptoms. Mild degrees of hyperthyroidism are acceptable as long as the patient tolerates this condition and pregnancy progresses satisfactorily. Patients should be followed closely, with careful monitoring of weight gain, heart rate, etc. A typical starting dose of ATD is 50-100 mg PTU twice daily (or the equivalent dosage for MMI/CMI). When both types of ATD are available, most clinicians prefer to use PTU than MMI. Monitoring ATD treatment monthly is necessary. After control of thyrotoxicosis, it is often possible to reduce the dose of ATD. The decision as to the maximal safe ATD dose during pregnancy is somewhat arbitrary, since the results vary with individual maternal-fetal pairs. Limits as widely variable as 150-450 mg/day have been suggested for PTU, but are of little use in practice.

Propranolol may be used transiently to control symptoms of acute hyperthyroid disease and for pre-operative preparation, and there are no significant teratogenic effects of propranolol reported in humans or in animals [315]. If a patient requires long-term propranolol administration, careful monitoring of fetal growth is advised, because of a possible association with intrauterine growth restriction [316].

Chronic use of iodide during pregnancy has been associated with neonatal goiter and hypothyroidism, sometimes resulting in asphyxiation because of tracheal obstruction [308, 317]. One study investigating the use of potassium iodide (6-40 mg/d), administered to selected hyperthyroid pregnant patients, showed improved maternal function with a normal neonatal outcome [318]. Because the experience with iodides is limited, iodide should not be used as a first line therapy for Graves’ disease during pregnancy, but they can be used temporarily, when needed, in preparation for a thyroidectomy.

Radioactive iodine administration is contraindicated during pregnancy. In case of inadvertent radioiodine administration, the fetus is exposed to radiation from mother’s blood (approximately 0.5-1.0 Rad per mCi administered). Since fetal thyroid uptake of radioiodine commences after twelve weeks, exposure to maternal radioiodine prior to this time is not associated with fetal thyroid dysfunction [319]. However, treatment with radioiodine after twelve weeks leads to significant radiation to the fetal thyroid. Multiple incidents of inadvertent exposure to radioiodine have been reported, causing fetal thyroid destruction, in utero hypothyroidism and subsequent neural damage [320, 321].

Thyroidectomy for Graves’ disease during pregnancy should be reserved as a second line of treatment in specific situations such as: a) persistent high ATD doses required to control maternal thyrotoxicosis; b) patients who present serious side effects to ATD (allergy, intolerance, etc.); c) non compliant patients; and finally d) rare cases with upper respiratory compressive symptoms due to goiter size. Thyrotoxic pregnant women should be prepared for surgery by using beta-blocking agents and a 10-14 days course of super-saturated potassium iodide solution (50-100 mg/d) in order to reduce vascularity of the thyroid gland. Surgery in pregnancy is deemed safest if it can be undertaken in the second trimester when organogenesis is complete, and thus the fetus is at minimal risk for teratogenic effects of medications, and the uterus is relatively resistant to contraction-stimulating eventsdrugss the fetus should be performed during the second trimester of gestation [248, 322].

The question of the safety of lactation during ATD therapy arises frequently [305]. Historically, women have been advised against breastfeeding when receiving ATD because of the presumption that ATD was present in breast milk in concentrations sufficient to affect infant’s thyroid function. Both PTU and MMI are secreted in human milk, although PTU less so because of more extensive binding to albumin [323-326]. In one study evaluating the effects of CMI (15 mg/d) or PTU (150 mg/d) on infants of nursing mothers, there was no evidence of neonatal hypothyroidism in the first weeks of life [327]. In another study, serum MMI levels were measured in breastfed infants of thyrotoxic mothers receiving MMI (20-30 mg/d): two hours after MMI ingestion, serum MMI levels in the babies were extremely low, far below the therapeutic range [328]. Thus, with both PTU and MMI, only limited quantities of the drugs are concentrated into milk. As long as the doses of MMI or PTU can be kept moderate (MMI: <20 mg/d; PTU: <250-300 mg/d), the risk for the infant is practically negligible and there is no evidence-based argument to advise mothers against nursing when they take ATD [285, 305, 329]. The drug should be taken by the mother after a feeding and it is prudent to monitor periodically the infant's thyroid function during the time of ATD administration to the mother, although a recent reassuring study showed that thyroid function in breastfed infants was not affected, even when ATD induced maternal hypothyroidism [330]. There is also a possibility that allergic reactions associated with ATD (agranulocytosis or rash) may occur in the infant. While these side effects are rare, they should be kept in mind when evaluating a febrile infant or presence of rash. In summary, within the limitations outlined above, the use of ATD in lactating mothers does not pose a risk to the neonate and appears to be safe.

TRAb may have stimulatory or inhibiting effects on the thyroid gland [331]. Thus, depending upon the balance between these opposite effects, TRAb production in pregnant women with Graves’ disease may stimulate or inhibit the fetal thyroid gland. Inhibitory TRAb production has been shown to cause hypothyroidism transiently in neonates born to mothers with Graves’ disease [150, 332].

When maternal TRAb production, with stimulating activity, is elevated and the antibody titers do not substantially decrease in the second part of gestation, fetal (and neonatal) hyperthyroidism constitutes a real risk [ 296-298, 333]. One to 5% of neonates of mothers with Graves’ disease have hyperthyroidism or neonatal Graves’ due to the trans-placental passage of stimulating maternal TRAb. The overall incidence is low because of the balance between stimulatory and inhibitory antibodies, and also ATD treatment of the mothers [296]. The incidence of neonatal Graves’ disease is not directly related to maternal thyroid function. Risk factors for neonatal thyroid dysfunction include history of a previously affected baby, prior radioiodine ablative treatment, and elevated TRAb titers at delivery [295, 296, 298, 334]. In the study by Mitsuda et al. for instance, the incidence of neonatal thyroid dysfunction was 67% if maternal TRAb was >130% and as high as 83% if TRAb was >150%, compared with only 11% if TRAb was <130% of normal (normal TRAb in this study was <115%) [295].

The risk of fetal hyperthyroidism can be assessed by fetal ultrasonographic data, showing the presence of fetal goiter, tachycardia, growth retardation, increased fetal motility, and accelerated bone maturation [286] ,[335]. If fetal hyperthyroidism is present, it is reasonable to initiate ATD treatment in the mother (PTU 200-400 mg or MMI 20 mg MMI), with thyroxine supplementation to maintain maternal euthyroidism, when needed. ATD treatment may thus be given on the presumptive diagnosis, but the definitive diagnosis of fetal hyperthyroidism may require umbilical cord blood sampling for fetal thyroid hormone determination. This procedure carries, however, a significant risk for the fetus (complications and/or fetal loss in 1% of cases) [336-340].

The administration of ATD to treat maternal Graves’ disease may cause fetal hypothyroidism, which should obviously be avoided in view of its potentially deleterious consequences on neuro-psychointellectual development. In clinical practice, the best way to avoid fetal hypothyroidism is to keep maternal circulating thyroid hormone levels in the upper quartile of the normality range [314, 341, 342]. In exceptional cases, radioactive iodine has unintentionally been administered to women with Graves’ disease who were unaware that they were pregnant and decided to maintain the pregnancy. In one such recently published case, the authors showed the advantage – despite the obvious risks associated therewith – of performing cordocentesis to predict the fetal outcome [343].

Subclinical hyperthyroidism (defined as a serum TSH below normal limits with free T4 and total T4 levels in the normal pregnancy range, and unaccompanied by specific clinical evidence of hyperthyroidism) can be seen in maternal Graves’ disease, but is more often found in association with gestational transient thyrotoxicosis (with/without hyperemesis gravidarum). Treatment of maternal subclinical hyperthyroidism has not been found to improve pregnancy outcome and may risk unnecessary exposure of the fetus to antithyroid drugs [336, 342, 344-346].

Fetal goiter may be associated with the treatment of maternal hyperthyroidism using ATD, result directly from the placental transfer of growth-stimulating effects of maternal TRAb, and finally also from direct inhibitory effects of ATD on the fetal thyroid gland inducing fetal hypothyroidism [347-352]. This complex – but important – topic was recently revisited. Gallagher et al. showed that the spectrum of neonatal thyroid dysfunction in maternal Graves’ disease receiving ATD ranged from frank hypothyroidism (secondary to the exposure to PTU and maternal blocking TRAb) to neonatal Graves’ thyrotoxicosis (secondary to exposure to maternal stimulating TRAb) and that the prenatal diagnosis remained often extremely complex [353]. Recently, the group of Michel Polak (Paris) reported their experience on a group of seventy-two mothers with past or present Graves’ disease, investigated using clinical evaluation, TRAb measurements, and ultrasound evaluation of the fetal thyroid gland and bone age [352, 354]. In the thirty-one pregnancies with negative TRAb and mothers without ATD treatment, all infants were normal at birth. In the remaining forty-one pregnancies, thirty were associated with positive TRAb and/or ATD treatment: fetal thyroid ultrasounds were normal at 32 weeks gestation and there was no evidence of fetal thyroid dysfunction. Finally, the remaining eleven fetuses had a goiter, of which seven were hypothyroid and four hyperthyroid. Fetal hyperthyroidism was associated with high maternal TRAb titers, accelerated bone maturation, and fetal goiter. Fetal hypothyroidism was associated with low TRAb titers, high maternal ATD doses, maternal serum T4 levels in the low normal range, delayed bone maturation, and goiter. The authors recommended TRAb measurement in women with current or past Graves’ disease at the beginning of pregnancy and close observation of pregnancies with elevated TRAb or ATD treatment by performing monthly fetal ultrasonography after 20 weeks of gestation.

Both thyrotoxicosis by itself and the administration of ATD to the expecting mother may raise concern with regard to potential teratogenicity of the disease and/or the medications. To date, it remains uncertain whether untreated Graves’ disease is associated with a higher frequency of congenital abnormalities [295, 305]. There have been reports of two distinct teratogenic patterns associated with MMI administration, namely aplasia cutis congenita and choanal/esophageal atresia, but data supporting these associations remain controversial [355-358]. Aplasia cutis (i.e. the absence of skin and accessory structures, usually over the scalp) has, so far, not been reported in mothers who received PTU. The so-called ‘methimazole embryopathy’ includes congenital abnormalities such as choanal and/or esophageal atresia, minor dysmorphic features and developmental delay [285]. After reviewing the evidence linking ATD treatment with such fetal abnormalities, one comes to the conclusion that the prevalence of these malformations is extremely rare: 2/241 MMI-exposed infants for instance in the series reported by Di Gianantonio [359]. In view of the potential danger – both for mother and offspring – of not treating active Graves’ disease in pregnancy, the concern posed by these rare congenital abnormalities would not, in our opinion, justify withholding ATD administration. However, despite the fact that the link between severe congenital defects and MMI exposure during pregnancy has not been formally established, it appears prudent to avoid the use of MMI during embryogenesis if PTU is available.

Undetected fetal thyrotoxicosis may be followed by thyrotoxicosis at birth. Neonatal thyrotoxicosis is considered to be uncommon, occurring in ~1% of pregnancies in patients with Graves’ disease in North America [5, 150]. In a recent reevaluation of this topic, it was suggested that the actual prevalence of neonatal thyrotoxicosis may be significantly higher, in the order perhaps of 2-10% [285, 296]. The risk appears highest in the offspring of women with not-well-controlled Graves’ disease, as well as in those with the highest TRAb titers. Mothers with a prior history of bearing infants with neonatal Graves’ disease are also at high risk of repeated episodes [150, 153]. Neonatal Graves’ disease is usually diagnosed at or shortly following birth, after maternal ATD has been cleared from neonatal serum and thyroid gland. Signs of neonatal thyrotoxicosis include congestive heart failure, goiter, proptosis, jaundice, hyper-irritability, failure to thrive, and tachycardia. Once considered, the diagnosis should be readily confirmed. Cord serum free T4 and TSH determinations should be performed in all deliveries of mothers with a history of Graves’ disease. Treatment should be carried out in conjunction with the neonatalogist, and may include iodide, ATD, glucocorticoids, digoxin, and beta-adrenergic blocking agents, depending on the cardiovascular status. Neonatal hyperthyroidism may have a delayed onset in some infants, particularly those in whom both anti-TSH receptor blocking and stimulating antibodies coexist. Thus, the pediatrician should be alerted to measure serum free T4 if symptoms suggesting thyrotoxicosis appear during the first 6-8 weeks of life, even if cord serum results were normal, and especially when cord serum TSH was suppressed [305, 335, 339, 341, 360-362].

There have been several reports of infants with neonatal central hypothyroidism, born to mothers with uncontrolled hyperthyroidism due to Graves’ disease during pregnancy [285, 363, 364]. The explanation was based on the fact that high maternal serum T4 levels, during a prolonged period of time, crossed the placental barrier and suppressed fetal TSH by feedback on the pituitary gland. In most cases, the diagnosis was made at birth or shortly thereafter, on the basis of a low neonatal serum total T4 contrasting with an inappropriately low serum TSH. In the majority of these infants, there was a return to euthyroidism within a few weeks or months.

The thyroid stimulating activity of human chorionic gonadotropin (hCG), its stimulatory effects on maternal thyroid function and its causal relation with hyperemesis gravidarum (HG) have already been discussed in the first section of this chapter (see ‘Effects of hCG on thyroid function’ and ‘Hyperemesis Gravidarum’).

Non autoimmune hyperthyroidism or gestational transient thyrotoxicosis “GTT” is characterized by elevated serum free T4 and T3 levels, suppressed TSH, variable clinical evidence of hyperthyroidism, usually minimal thyroid enlargement, and absence of thyroid auto-antibodies. The syndrome occurs transiently near the end of the first trimester of gestation, usually in hitherto healthy women who have otherwise a normal pregnancy, and it is frequently associated with excessive vomiting [6, 14, 39, 41, 42, 329, 374]. GTT differs fundamentally from Graves’ disease in that it occurs in women without past history of Graves’ disease and absence of detectable anti-TSH receptor antibodies (TRAb). GTT is not always clinically apparent, due to its transient nature. Its etiology is related to the thyrotropic stimulation of the maternal thyroid gland associated with elevated hCG levels (see Figure 14-17).

The prevalence of GTT is highly variable in populations of normal pregnancies. Prospective studies carried out in Europe have indicated that the prevalence may reach 2-3% of unselected pregnancies (that is 10-fold more prevalent than Graves’ disease) [194]. In other regions of the world, the prevalence of GTT appears to be as low as 0.3% (Japan) or as high as 11% (Hong Kong) [375, 376].

Figure 17. Human CG levels measured at different time points during gestation in women with a diagnosis of GTT. The curve flanked by the dashed lines in the lower part of the graph represents the mean hCG levels (with 95% confidence intervals) determined in a cohort of normal pregnancies, showing the classical hCG peak near the end of the first trimester. The individual points show serum hCG levels in fifteen women with GTT, at the time of their recall, i.e. 3-6 weeks after initial screening for thyroid function. Initial diagnosis was based on suppressed serum TSH with elevated free T4 concentrations. All women clearly had abnormally elevated hCG values, even when the measurements took place several weeks after peak hCG. In a few cases, hCG was measured again during follow-up in the 2nd trimester and the results showed sustained abnormally elevated serum hCG levels. (from Glinoer, Ref 32)

Owing to the transient nature of the syndrome, clinical manifestations of hyperthyroidism are not always apparent or routinely detected. When women were specifically questioned about possible symptoms compatible with thyrotoxicosis, we found weight loss (or absence of weight increase), tachycardia and unexplained fatigue in one half of the women with GTT. Hyperemesis gravidarum was frequently associated with the most severe cases, and in a few women symptoms were sufficiently alarming to justify hospitalization for treatment. In cases followed until parturition, GTT was always transient; elevated serum free T4 values reverted gradually to normal in parallel with the decrease in hCG concentrations. Serum TSH often remained partially (or totally) suppressed for several weeks after free T4 reverted to normal, i.e. until after mid-gestation [377, 378]. GTT was not associated with a less favorable outcome of pregnancy. It should also be noted that GTT may occur, by coincidence, in women with preexisting thyroid disorders, such as glandular autonomy, autoimmune thyroiditis or Graves’ disease, and even in women with genetic resistance to thyroid hormone [379]. The association of GTT with an underlying thyroid abnormality often leads to more severe presentations of thyrotoxicosis. The stimulatory effect of hCG may also help explain the exacerbation of thyrotoxicosis due to Graves’ disease, that is sometimes encountered during the 1st trimester. Finally, when women with GTT are followed in subsequent pregnancies, the syndrome has a characteristic tendency to relapse.

Normal women develop GTT when they have abnormally elevated peak hCG levels, and when high hCG values are sustained during an unusually prolonged period [380]. Twin pregnancy is a naturally occurring clinical condition associated with sustained and high hCG concentrations (see Figure 14-18). Peak hCG values were shown to be significantly higher (almost double) and of a much longer duration in women with a twin pregnancy. While peak hCG values lasted only for a few days in singleton pregnancy, peak hCG levels (>100,000 UI/L) lasted for up to six weeks in twin pregnancies. Twin pregnancy was associated with a more profound and 3-fold more frequent suppression of serum TSH values. Also, while serum free T4 levels remained unaltered in singleton pregnancy, they often rose transiently above normal in twin pregnancy. Symptoms related to thyrotoxicosis were usually mild or absent, except for more intense vomiting which was more frequently noted (obstetricians often observe increased nausea and vomiting in women with a twin pregnancy).

Figure 18. Profiles of changes in hetero-dimeric intact hCG, serum free T4 and serum TSH, comparing women with singleton (pale blue) and thirteen women with a twin pregnancy (pink color). Each point corresponds to the mean (± 1 sd) of individual serum samples measured at each gestational age (from Grün, Ref 380).

No data are available to delineate precisely which patients with GTT require a specific treatment. In most cases, no specific treatment is required and the symptoms can be relieved by the administration of beta-adrenergic blocking agents for a short period, while waiting for spontaneous recovery of elevated thyroid hormones to occur. In patients with a severe clinical presentation (clearly symptomatic hyperthyroidism), some – but not all – authors suggest giving treatment with PTU, usually for a few weeks only, and therapy is often discontinued by mid-gestation. As already indicated in the section on ‘treatment for Graves ‘disease’, treatment of subclinical hyperthyroidism has not been found to improve pregnancy outcome and may risk unnecessary exposure of the fetus to ATD [336, 342, 344-346].

The precise pathogenic mechanisms underlying GTT are still not fully understood but the etiology of the syndrome is felt to be hCG itself or derivatives of hCG [14]. Based on the example of GTT associated with twin pregnancy, a quantitative direct effect of elevated hCG levels to stimulate the thyroid gland is presumably sufficient to explain hyperthyroidism in most pregnant women, provided that hCG values remain above 75,000-100,000 UI/L for a sufficient period of time: GTT is directly related to both the amplitude and duration of peak hCG values [380]. Human CG acts as a weak TSH agonist, increasing cAMP production, iodide transport and cell growth in thyrocytes [42, 49, 50]. It remains possible that abnormal h CG molecular variants, with a prolonged half life, are produced in these situations explaining sustained prolonged high circulating hCG levels [30]. It has also been hypothesized that a dysregulation of beta-hCG production may transiently take place in these women [24]. Finally, hCG molecular variants with a more potent thyrotropic activity have been detected, although these variants appear to be more specifically found in women with hydatidiform mole or choriocarcinoma [381-383].

Whatever the final explanation, the effects of hCG to stimulate the thyroid gland can best be explained by the marked homology that exists between the hCG and TSH molecules, as well as between the LH/CG and TSH receptors [384, 385]. Gestational non autoimmune hyperthyroidism can be considered an example of an endocrine ‘spill-over’ syndrome, a concept based on molecular mimicry between hormone ligands and their receptors [14, 50, 281, 381].

An interesting question – although unresolved – is whether the thyrocyte is a passive bystander of abnormal thyrotropic activity of hCG in GTT, or whether the thyrocyte itself, through variable degrees of sensitivity of the TSH receptor, may play an active role in its responsiveness to hCG’s stimulatory effect. A woman with recurrent gestational hyperthyroidism was recently reported by Rodien et al [386]. After two miscarriages, she presented hyperemesis early in pregnancy with overt hyperthyroidism, requiring treatment with PTU. During her next pregnancy, she experienced a relapse of the same situation. The patient’s mother had also been diagnosed with hyperthyroidism during her 2nd and 3rd gestations, mistakingly considered to be Graves’ disease. Study of the TSH receptor of the patient disclosed a single mutation in the extracellular domain of the TSH receptor (K183R), rendering this mutant receptor highly sensitive to hCG, and accounting for recurrent thyrotoxicosis during pregnancies in the presence of normal hCG levels. Thus so far, one fascinating example has been reported of a substantially increased sensitivity of the TSH receptor to the stimulatory effect of hCG in humans. This unique finding raises the possibility that some women who develop GTT may have an abnormality at the level of the thyroid follicular cell (see Figure 14-19).

Figure 19. TSH receptor mutation, with a Lysine to Arginine mutation in position 183 of the ecto-doamin. The graph on the left shows that the mutation confers high sensitivity to hCG (red curve), compared with wild type TSH receptor (blue curve). The family tree (upper right) shows the pedigree of the patient. (from Rodien & Vassart, Refs 386-388).

In further studies of structure-phenotype relationship at the level of the TSH receptor, the group of Vassart carried out site-directed mutagenesis, substituting lysine 183 in the ecto-domain of the TSH receptor by a variety of amino acids expressing different physicochemical properties [16, 387]. Somewhat unexpectedly, all mutant TSH receptors displayed a widening of their specificity toward hCG stimulation. Modeling of the mutated receptors indicated that the increased gain of sensitivity might result from the release of a nearby glutamate residue (in position E157) from a salt bridge formed with K183. As of now, this situation remains exceptional since most patients with gestational non autoimmune transient hyperthyroidism do not seem to be familial and almost invariably have hCG levels >100.000 UI/L [388].

As already alluded to in the first sections of this Chapter, GTT is often associated with nausea (morning sickness), increased vomiting and hyperemesis gravidarum, a severe condition requiring hospitalization and drastic treatment. Several studies have established a correlation between the severity of vomiting and abnormal thyroid function. Thirty to sixty percent of patients with hyperemesis gravidarum have elevations of free thyroid hormone concentrations, along with a suppressed TSH, and in the absence of any evidence of Graves’ disease [38, 42, 389]. A common observation among obstetricians is also that women with a twin pregnancy often experience more severe vomiting during early gestation. Increased vomiting is not usually associated with hyperthyroidism due to Graves’ disease in pregnancy. Thus, hyperemesis appears to be mainly associated mainly with hCG-induced thyrotoxicosis, although evidently all women who vomit during early pregnancy do not have disturbances of thyroid function. The most likely explanation is that elevated and sustained hCG levels in the circulation promote estradiol production in these women; the combination of elevated hCG, estradiol and free T4 concentrations then transiently promotes emesis near the period of peak hCG, by some yet not fully understood mechanism [ 14, 42, 50, 376, 381]. A recent matched-paired study in fifty-eight pregnant Chinese women showed that women with hyperemesis gravidarum were younger and had higher serum free T3 and T4 levels, as well as higher β-hCG concentrations [390]. In this study, a logistic regression analysis of the data showed that the principal determinant for having hyperemesis was high the free T4, and not the high hCG levels. Thus, these results reinforce the concept that hCG is not independently involved in women with hyperemesis gravidarum, but rather indirectly by stimulating the thyroid to induce gestational hyperthyroidism.