To determine whether sonography can be used to distinguish hyperthyroidism from hypothyroidism in pregnancies with fetal goiter.
To determine whether sonography can be used to distinguish hyperthyroidism from hypothyroidism in pregnancies with fetal goiter.
This was a retrospective study of 39 cases of fetal goiter. The majority of the mothers had Graves' disease. Fetuses were scanned for the existence of a hypertrophic thyroid gland (goiter) beginning at 22 gestational weeks. Once a goiter was diagnosed, different echographic features were analyzed and the effect of chosen treatment on fetal thyroid development was monitored.
On color Doppler, 68.8% of hypothyroid goiters had a peripheral vascular pattern vs. 20% in cases of fetal hyperthyroidism (P = 0.0574). No hypothyroid goiter presented central vascularization whereas half the hyperthyroid goiters did (P = 0.0013). Fetal tachycardia was a good indicator of hyperthyroidism (57.1% v.s 6.3%; P = 0.0055). Delayed bone maturation was seen in hypothyroid goiters (46.9% vs. 0%; P = 0.0307), while advanced bone maturity was specific to hyperthyroid goiters (85.7% vs. 0%; P < 0.0001). Lastly, an increase in fetal movement was observed in cases of fetal hypothyroidism (43.8% vs. 0%; P = 0.0364).
Based on the color Doppler pattern of goiter, fetal heart rate, bone maturation and fetal mobility, we established an ultrasound score to predict fetal thyroid function in cases of fetal goiter. Copyright © 2009 ISUOG. Published by John Wiley & Sons, Ltd.
Fetal goiter, which signals a thyroid dysfunction, is a rare condition: the incidence of goitrous hypothyroidism is 1/30 000–50 000 live births1, while that of goitrous hyperthyroidism is unknown. Fetal goiter can cause complications attributable either to the goiter itself, such as a voluminous cervical mass (with polyhydramnios due to compression of the upper airways, anasarca or heart failure) or to the thyroid dysfunction. Fetal hyperthyroidism may cause intrauterine growth restriction (IUGR) with accelerated bone maturation, intrauterine death by cardiac failure or thyrotoxicosis, craniosynostosis with intellectual impairment2, or neonatal thyrotoxicosis3. Fetal hypothyroidism can delay bone maturation and psychomotor development4–8.
When the mother has Graves' disease, hypertrophy of the fetal thyroid gland may arise because of iatrogenic fetal hypothyroidism caused by transplacental passage of antithyroid drugs, or because of fetal hyperthyroidism due to transplacental passage of antibodies against the thyroid-stimulating hormone (TSH) receptor (TRAb) (the occurrence of inhibiting antibodies being exceptional). In these women, absence of fetal goiter or thyroid hypertrophy has been described as a sign of immediate fetal and neonatal euthyroidism, while development of fetal goiter in utero is taken to indicate an underlying thyroid dysfunction9.
Assay of thyroid hormones in the amniotic fluid does not strictly reflect fetal thyroid status, and the relative contributions of fetus and mother are unknown10–13. At present, fetal thyroid status can only be determined with certainty by fetal blood sampling (FBS), but this has a complication rate of approximately 1%14–17.
We therefore aimed, in an attempt to avoid invasive methods, to investigate the use of ultrasound in distinguishing fetal hyper- from hypothyroidism. On the basis of our observations of a series of 39 cases of fetal goiter, mainly in women with Graves' disease, we now propose an ultrasound score predictive of fetal thyroid status.
This was a prospective study of 39 cases of fetal goiter managed at the Robert Debré Teaching Hospital in Paris, France from 1993 to 2006. Cases were identified by screening because of maternal hyperthyroidism (mostly Graves' disease), or fortuitously when there was no history of thyroid dysfunction in the mother.
Cases were managed following our published protocol9. Once a month, starting at 22 weeks of gestation, fetal heart rate was recorded, and a fetal ultrasound examination was performed for measurements of thyroid gland size (diameter and circumference), determination of fetal growth parameters and evaluation of fetal bone maturation. Except for the first 10 cases, which were examined with an 128XP/10 Acuson ultrasound machine (Acuson, Moutain View, CA, USA) with variable focus and a 3.5-MHz sector transducer, fetal sonograms were perfomed using an EVB 525 variable focus ultrasound machine (Hitachi, Hialeah, FL, USA) with a 3.5-MHz sector transducer. The same tuning and adjustments were applied to each device.
Fetal goiter was defined as a thyroid circumference or diameter >95th percentile for gestational age, according to the normative data established by Ranzini et al.18 and to our own unpublished nomogram established by one of the authors (E.V.) in 250 patients, which is consistent with that of Ranzini et al.. When fetal goiter was found, color flow Doppler examination of the thyroid was performed with the velocity scale adjusted to 13 cm/s, as described previously9, 19, 12, to determine whether there was no, peripheral or central vascularization (Figures 1 and 2). Peripheral vascularization is thought to reflect trophic vascularization of a hypertrophic but inactive thyroid gland, whereas central vascularization reveals global vascularization of an overactive thyroid gland, as in Graves' disease.
Bone maturation was evaluated on the sonogram obtained at 32 weeks. Normally, before 28 weeks the distal femoral ossification center is undetectable, around 32 weeks it is dot-like in appearance and not visible consistently, before 33 weeks it is <3 mm in diameter and not visible consistently, and after 35 weeks it is 7–8 mm in diameter and visible consistently. Accelerated bone maturation was defined as the presence of the distal femoral ossification center before 31 weeks (Figure 3), and delayed bone maturation was defined as the absence of this center after 33 weeks20.
Fetal tachycardia was defined as a fetal heart rate continuously faster than 160 bpm.
After multidisciplinary evaluation and discussion, fetal hyperthyroidism was treated by increasing antithyroid drug dosage, while fetal hypothyroidism was treated by reducing antithyroid drug dosage and, if needed, by administration of intra-amniotic L-thyroxine9, 21. Treatment was decided according to fetal thyroid status, as indicated by clinical, ultrasound and laboratory findings. Prenatal diagnosis of fetal status was validated if ultrasound criteria improved after initiation of treatment. In difficult cases, FBS was performed, and the choice of treatment was checked at birth by means of cordocentesis. TSH, free triiodothyronine and free thyroxine were measured using chemiluminescence immunoassay with the ACS-180SE system (Bayer Diagnostics, Westwood, CA, USA). The values were interpreted according to gestational or postnatal age22.
Statistical analysis was performed using Student's t-test (GraphPad Prism, version 4, Windows, Microsoft Corp., Redmond, WA, USA), or χ2 or Fisher's exact test (SAS, version 8.2, SAS Institute, Cary, NC, USA) if the constraints of Student's t-test were not respected. Values are given as mean ± SD or percentage.
Population characteristics are summarized in Table 1. The study population included 38 pregnancies, one of which was a twin pregnancy in which both fetuses presented with a goiter (Cases 8 and 8a), giving 39 fetuses. There were 37 mothers, one of whom was included in the study for two successive pregnancies, in both of which fetal goiter was diagnosed (Cases 1 and 2).
|Case||Maternal characteristics||Fetal characteristics|
|Maternal thyroid disease||Antithyroid drugs*||TRAb maximum†||Maternal thyroid status‡||Fetal thyroid status§||GA (weeks)¶||FBS (n)||Cord blood sampling at delivery|
|1||Graves' disease||Yes||21N||Hypo.||Hyper.||23.5||2||Euthyroid (TSH unmeasurable)|
|2||Graves' disease||Yes||78N||Hypo.||Hyper.||29||1||Euthyroid (TSH unmeasurable)|
|5||Graves' disease||Yes||2.8N||Euthyroid**||Hypo.||38||0||Compensated hypo.|
|8||Graves' disease||No||Negative||Hypo.||Hypo.||35||0||Compensated hypo.|
|8a||Graves' disease||No||Negative||Hypo.||Hypo.||36||0||Compensated hypo.|
|9||Graves' disease||Yes||8N||Hypo.||Hypo.||38||0||Compensated hypo.|
|10||Graves' disease||Yes||Negative||Euthyroid||Hypo.||36||0||Compensated hypo.|
|11||Toxic nodular goiter||Yes||Negative||Euthyroid||Hypo.||29||1||Euthyroid|
|12||Graves' disease||Yes||Negative||Euthyroid**||Hypo.||23||0||Stillborn (hypo.)|
|16||Graves' disease||Yes||1.5N||Euthyroid**||Hypo.||32||1||Hypo. changed to hyper.|
|18||Graves' disease||Yes||52N||Euthyroid||Hyper.||22.5||1||Intrauterine death (TSH unmeasurable)|
|20||Graves' disease||No||24N||Euthyroid||Hyper.||32.5||1||Euthyroid (TSH unmeasurable)|
|25||Hyper. in pregnancy||Yes||Negative||Hypo.||Hypo.||26||1||Euthyroid|
|34||Graves' disease||Yes||6.2N||Euthyroid||Hypo.||25.5||1||Compensated hypo.|
|36||Graves' disease||Yes||17.2N||Hypo.||Hypo.||22||2||Compensated hypo.|
|37||Graves' disease||Yes||5.5N||Euthyroid||Hypo.||28||1||Compensated hypo.|
Thirty-five mothers had a history of past or current thyroid dysfunction, of whom 31 had confirmed Graves' disease, which in one case progressed to Hashimoto's disease (Case 26). Two mothers presented with unclassified hyperthyroidism (Cases 24 and 38). One mother presented with a nodular goiter diagnosed at the start of pregnancy (Case 11). In another mother (Case 25), uncontrollable vomiting in early pregnancy led to a diagnosis of transient hyperthyroidism, and TRAb remained negative throughout the pregnancy. Of the two mothers with no history of thyroid disease, one had been treated with potassium iodide between 16 and 21 weeks of gestation, because of nodular erythema (Case 22)26; in the other case (Case 6), the neonate was later diagnosed with hypothyroidism, linked to altered hormone synthesis.
Twenty-two mothers were euthyroid when fetal goiter was diagnosed. Of these, 10 had free thyroxine in the lower third of the normal laboratory range (Case 3, 4, 5, 12, 13, 16, 23, 26, 28 and 35). For 15 of the affected fetuses, the mother was hypothyroid when fetal goiter was diagnosed, 12 of whom (including the mother with two successive pregnancies) were likely to have received a large dose of antithyroid drug treatment. Two mothers were hyperthyroid (Cases 30 and 31) in spite of antithyroid drug treatment.
There were seven hyperthyroid fetuses. Four were in euthyroid mothers, two were different pregnancies in one hypothyroid mother (Cases 1 and 2), and in the other the mother was hyperthyroid.
The two mothers with a history of, or presenting with, Graves' disease who were not treated with antithyroid drugs, and did not have positive TRAb during pregnancy, (Cases 8/8a and 26) had elevated antithyroid peroxidase antibody (3750 U/mL at 4 months postpartum in Case 8/8a and 1240 U/mL during pregnancy in Case 26).
Fetal goiter, or hypertrophy, was detected by ultrasound examination at an average of 29 (range, 22 to 38) weeks of gestation. FBS was performed in 24 cases, and twice in six of these. A second FBS was done when ultrasound monitoring showed no improvement in the fetal thyroid status, despite intra-amniotic injections of levothyroxine to treat hypothyroid fetuses, or high doses of propylthiouracile to treat hyperthyroid fetuses and which resulted in maternal hypothyroidism (that was subsequently treated with levothyroxine).
Eighteen of the 32 (56.3%) cases of fetal goitrous hypothyroidism were confirmed by FBS. In one case (Case 22), in which the fetus was exposed to potassium iodide in utero, goitrous hypothyroidism was suspected, but FBS indicated euthyroidism. We diagnosed transient goitrous hypothyroidism that resolved spontaneously, as FBS was carried out after natural correction of fetal thyroid status. In Case 12, goitrous hypothyroidism was diagnosed antenatally. FBS was scheduled, but premature delivery occurred (26 + 4 weeks) related to infection, in a setting of severe metrorrhagia related to placenta previa, and the child was stillborn.
Six of the seven (85.7%) cases of goitrous hyperthyroidism were confirmed by FBS. In the seventh case (Case 3), FBS did not confirm goitrous hyperthyroidism, and at birth the child presented with iatrogenic hypothyroidism due to an excessive increase in propylthiouracil. However, accelerated closure of the fontanelles observed at birth confirmed in utero hyperthyroidism.
Table 2 summarizes the ultrasound findings for hypo- and hyperthyroidism. No case of polyhydramnios or hydrops was observed. However, excess amniotic fluid was found in four cases of hypothyroidism (Cases 13, 14, 17 and 25) and in three cases of hyperthyroidism (Cases 1, 2 and 19). There was one case of moderate tracheal compression (Case 25) and one case of cardiac sequelae, with cardiomegaly and high cardiac velocity, in a setting of hyperthyroidism (Case 18). Malformation was seen in only one case of aplasia cutis congenita (Case 13) caused by use of carbimazole in early pregnancy.
|Fetal thyroid status||Vascularization of goiter||Tachycardia||Bone maturation||Fetal movements increased|
|Hypothyroidism||22/32 (68.8)||0||2/32 (6.3)||15/32 (46.9)||0||14/32 (43.8)|
|Hyperthyroidism||1/5 (20)||3/5 (60)||4/7 (57.1)||0||6/7 (85.7)||0|
Color Doppler was not performed in one of the earliest cases (Case 1), before its systematic application. It was recorded as positive in Case 3, but no further details were given, and review of the ultrasound images did not allow a posteriori interpretation. Color Doppler revealed peripheral vascularization (Figure 1) in just one of the five documented hyperthyroid goiters and in 22 of the 32 (68.8%) hypothyroid goiters. It revealed central vascularization (Figure 2) in three of the hyperthyroid goiters and was not seen in any hypothyroid goiter.
Six fetuses presented with permanent tachycardia (>160 bpm), which was associated with hyperthyroidism in four. In Case 12, one of the two cases of fetal hypothyroidism and tachycardia, the infant was stillborn at 26 + 4 weeks, following failure of tocolysis, in a setting of severe metrorrhagia, low-lying placenta and infection. A normal heart rhythm was seen in three fetuses with hyperthyroidism. Permanent slowing of heart rhythm was not seen in any fetus.
Bone maturation (Figure 3) was delayed in 15 fetuses, all of which presented hypothyroidism, and was accelerated in six fetuses, all with hyperthyroidism.
In adults, hyperactive behavior is associated with hyperthyroidism, while psychomotor slowdown suggests hypothyroidism. We therefore paid particular attention to fetal movements. Surprisingly, increased fetal movements were seen in 14 fetuses with hypothyroidism, but in no fetus with hyperthyroidism. Of note, the hyperthyroid fetuses had normal movement patterns whereas in the majority of hypothyroid fetuses we observed jerky movements. We recognize that these data are subjective, but they derive from a scientific approach, with careful, prospective observation. For now, a more precise definition cannot be given.
Although the four abovementioned sonographic signs are not specific, it seems that hypothyroidism and hyperthyroidism give appreciably different ultrasound findings for goiter vascularization, fetal heart rate, bone maturation and fetal movements, and that Doppler examintation of the fetal goiter could guide diagnosis. We therefore propose a diagnostic ultrasound score for use in cases of fetal goiter (Table 3). This score is formed by scoring each of the four signs, and adding the individual scores together to give an overall score, ranging from −1 to 4, in which a score ≥ 2 is suggestive of hyperthyroidism and a score <2 is indicative of hypothyroidism (Table 4 and Figure 4). To establish the score, we excluded cases with incomplete ultrasound findings (Cases 1 and 3), although these cases would have had scores of 3 or 4 and 2 or 3, respectively, on the basis of their Doppler data, correctly placing them both in the category of hyperthyroidism. We also chose not to take into account the 26.5-week stillbirth in an infectious setting (Case 12), as diseases impacting on fetal well-being may affect the score components, resulting in the risk of misclassification when using the score. Its score would have been 2 with tachycardia. Our new score correctly classified all 36 remaining fetuses as hypo- or hyperthyroid.
|Peripheral or absent||0|
|Fetal heart rate|
|Case||Vascularization||Fetal heart rate||Bone maturation||Fetal movements||Score||Fetal thyroid status|
To assess a posteriori our prenatal management, we examined neonatal thyroid function. Cord blood sampling showed normal triiodothyronine and thyroxine values in 22 of 26 (84.6%) neonates whose goiter was shown antenatally to be associated with hypothyroidism. Case 16 showed unusual values, with elevated triiodothyronine and thyroxine but also elevated TSH, as if fetal thyroid function was switching from hypothyroidism to hyperthyroidism after interruption of propylthiouracil treatment. Five of the seven (71.4%) fetuses with hyperthyroidism had normal triiodothyronine and thyroxine values. The fetus that died in utero was considered as a treatment failure, despite having normal triiodothyronine and thyroxine values.
We have previously found fetal goiter in 2.6% of pregnancies in which the mother had a history of thyroid disease, and in 19% of fetuses carried by mothers with Graves' disease21. Ultrasound examinations in such cases should be performed monthly from 22 weeks of gestation if TRAb are present and/or if the mother is on antithyroid drugs27. Some authors have recommended ultrasound monitoring of the fetal thyroid when the mother is treated with propylthiouracil28, 29, while admitting the limitations of the method. Cohen et al.28 reported ultrasound detection of goiter in five fetuses, but, as the ability of ultrasound to distinguish between hyper- and hypothyroidism had not yet been proved, they resorted to FBS when the goiter did not regress despite antithyroid drug dose reduction. The solution to the problem of classification of fetal dysthyroidism seems to lie in refinement of the ultrasound examination and development of techniques such as magnetic resonance imaging (MRI).
Three-dimensional (3D) ultrasound could prove useful in the analysis of fetal goiters. In a case of fetal goiter examined by 3D ultrasound, the authors were unable to delay FBS, which the patient underwent twice30. However, they indicated that the 3D image facilitated the mother's understanding of fetal goiter, and therefore led to better treatment compliance. In 2003, MRI was used by Matsumoto et al.31 to observe goitrous hypothyroidism during pregnancy. A homogeneous mass was noted, with T1 signal hyperintensity. One year later, Kondoh et al.32 reported a case of goiter seen on MRI in a 32-week fetus with congenital hypothyroidism. The image obtained was of a goiter with T1 and T2 signal hyperintensity, in comparison with the signal from the muscle tissue. To our knowledge, there are as yet no published reports of signal intensity in antenatal hyperthyroidism although it is possible that MRI could, in time, help discriminate between hyper- and hypothyroidism.
In this study we investigated various sonographic signs, including bone maturation, fetal heart rate, fetal movements and vascularization. Bone maturation is delayed only in severe forms of hypothyroidism20, 33, 34 and can be detected in the majority of cases after 32 weeks of gestation. Physiologically, one might expect permanent slowing of fetal heart rate or decreased fetal movements to accompany fetal hypothyroidism, yet none of these signs has been reported in the literature, and they have never been observed by us. In fact, we observed increased fetal movements in cases of hypothyroidism. Fetal hyperthyroidism can be suspected when there is IUGR, sometimes accompanied by oligohydramnios (very low specificity)35, 36, accelerated bone maturation or fetal tachycardia (nonspecific and inconstant)35–38. Other signs accompanying goiter are not discriminant, such as polyhydramnios being related to the goiter volume rather than etiology6, 11, 29, 31, 33, 39–50. At an advanced stage of in utero dysthyroidism, even heart failure (with or without anasarca) is not unique to hyperthyroidism51 and may be seen in hypothyroidism45, 52. In short, all these signs are inconstant, non-specific or observed too late, and are of no value individually, yet combining some of them in our new score proved effective.
Regarding vascularization, formerly, an avascular anterior neck mass seen on antenatal ultrasound would generally have been identified as goiter, while a vascular mass would have been assumed to be angioma53. Later, a Doppler ‘flash’ was considered to indicate hyperthyroidism. Yet, while hypervascularization has been described in cases of hyperthyroidism19, 54, it may also be seen in hypothyroidism39, 45, 55, 56. In 1997, we reported normalization of Doppler flash in a fetus with hyperthyroidism, after initiation of propylthiouracil treatment in the mother19, showing that color Doppler can be used to assess the efficacy of antithyroid drugs in fetal hyperthyroidism. This was also true in a case of fetal hypothyroidism after intra-amniotic injection of thyroxine45. In 2000, we associated for the first time central Doppler flash with hyperthyroidism and peripheral flash or absence of flash with hypothyroidism. Although specificity was not absolute, disappearance or reduction of the Doppler flash seemed to be associated with improvement in fetal thyroid status, and as such seemed useful for monitoring21. Our ultrasound score, however, was established to facilitate primary diagnosis, and the score's relevance in monitoring of goiters is uncertain.
Increased fetal movements were suggestive of hypothyroidism. We do not have a clear explanation for this paradoxical sign, which is at odds with the clinical picture seen in adults. However, an early study reported jerky and rapid movements in 50% of Down syndrome fetuses57 and congenital hypothyroidism is not uncommon in Down syndrome newborns (1.8–2.9%)58, 59. It may be that high type II iodothyronine deiodinase activity in the brain, which is physiological in the fetus, is increased further in hypothyroidism60. This enzyme converts thyroxine to triiodothyronine, whose local overproduction in the brain could explain fetal hyperactivity. In contrast, in the adult, thyroxine-to-triiodothyronine conversion is carried out by type I iodothyronine deiodinase, whose activity is slowed in hypothyroidism60.
We observed peripheral vascularization in 68.8% of cases of hypothyroidism, and in a single case (20%) of hyperthyroidism. Central vascularization was seen only in cases of hyperthyroidism (60% of these cases). One potential problem during ultrasound examination would be the erroneous deduction of central vascularization from a Doppler signal increased artifactually by the exaggerated movements of a fetus with hypothyroidism, a pitfall rendered even more likely as the Doppler speed must be low (13 cm/s).
In summary, our ultrasound score can guide diagnosis in cases of fetal goiter. FBS could be used secondarily, if initial treatment fails. We established our score retrospectively and recently it has been used prospectively in our unit as a diagnostic tool. Further prospective study will reinforce its validity, but this score is readily applicable in at-risk pregnant patients with Graves' disease.