SUY GIÁP BẨM SINH

 Abstract

OBJECTIVE. The purpose of this study was to retrospectively evaluate the use of sonography as the primary imaging modality for congenital hypothyroidism (CH).

MATERIALS AND METHODS. From our regional registry, we reviewed the cases of patients for whom either sonography or ^99mTc-pertechnetate scanning was performed for CH between 2003 and 2010. Ultrasound studies were reviewed for presence, size, echotexture, vascularity, and location of the thyroid gland. Technetium-99m-pertechnetate scans were evaluated for the presence and location of the thyroid gland. The ultrasound studies were compared with the ^99mTc-pertechnetate scans. We assessed the use of ultrasound as the primary imaging modality for the evaluation of CH.

RESULTS. We identified the cases of 124 patients (89 girls, 35 boys). Ultrasound studies were available for 121 patients, and ^99mTc-pertechnetate studies for 62 patients. Three patients were examined only by ^99mTc-pertechnetate scanning. The final imaging results were normal location with normal size or diffuse enlargement of the thyroid gland (n = 47), sublingual thyroid gland (n = 49), agenesis (n = 18), hypoplasia (n = 8), and hemiagenesis (n = 2). Compared with ^99mTc-pertechnetate scanning, ultrasound had high (100%) specificity and low (44%) sensitivity for detection of sublingual thyroid gland.

CONCLUSION. We suggest using ultrasound as the primary imaging modality for guiding the treatment of children with CH, potentially decreasing radiation exposure and cost.

 

Congenital hypothyroidism (CH) is defined as thyroid hormone deficiency present at birth. It can be subdivided into permanent and transient types. Permanent CH refers to persistent deficiency of thyroid hormone that requires lifelong treatment [1]. Transient CH refers to a temporary deficiency of thyroid hormone. The deficiency is present at birth, but recovery to normal thyroid hormone production usually occurs within the first few months or years of life.

Almost all neonates are screened for CH. A heelstick is performed to evaluate the level of thyroid-stimulating hormone (TSH). All infants with high TSH levels are considered to have CH until proven otherwise [1]. Most cases of CH are asymptomatic at birth, but if left untreated, the condition can lead to growth failure and profound mental retardation between 3 and 6 months of age [1–5]. The incidence of CH is 1:3000–1:4000 [1, 6].

In undeveloped countries, the most common cause of CH is iodine deficiency (transient CH), but in the developed world, 85% of cases of CH are caused by thyroid dysgenesis (aplasia, hypoplasia, or ectopia). Inborn errors of thyroid hormone biosynthesis (dyshormonogenesis) or defects in peripheral thyroid hormone transport, metabolism, or action account for 10–15% of cases and are also associated with genetic defects. Secondary, or central, CH may occur with isolated TSH deficiency, but more commonly it is associated with congenital hypopituitarism [1].

Determining the cause of CH guides management and genetic consultation because it has prognostic implications [1]. Although thyroid hormone replacement is the initial treatment in all cases, if the patient has a normal-appearing eutopic thyroid gland, a trial of discontinuing levothyroxine when the patient is approximately 3 years old is often undertaken to differentiate permanent from transient CH. If the thyroid gland adequately functions, no further replacement hormone is required. If no thyroid tissue is found or if dyshormonogenesis has occurred, the child needs thyroid supplementation for life.

Imaging studies to help determine the underlying cause of CH include thyroid radionuclide examinations and thyroid ultrasound. Thyroid radionuclide studies with ^99mTc-pertechnetate or ¹²³I are considered the standard for imaging in the evaluation of thyroid dysgenesis. Although ^99mTc-pertechnetate is preferred because of lower thyroid and total body radiation dose (≈ 0.04 mSv compared with 0.35 mSv) [7], both result in radiation exposure to the patient. In the case of eutopic location of the thyroid gland, an ¹²³I uptake followed by a ^99mTc-pertechnetate perchlorate discharge test is the definitive study for identifying an organification defect of the thyroid gland [8].

Sonography does not involve the risk of ionizing radiation and can be used to differentiate thyroid dysgenesis and other causes of CH in which the thyroid gland has normal morphologic features [9, 10]. Sonography, however, has lower sensitivity than ^99mTc-pertechnetate scintigraphy in the detection of sublingual thyroid. The use of color Doppler ultrasound (CDUS), however, has been found to increase the detection of sublingual ectopic thyroid [1, 2, 5, 11].

For several years at our facility, we have been using sonography as the primary screening imaging modality in the care of patients with CH and using ^99mTc-pertechnetate scintigraphy primarily for patients with thyroid dysgenesis. In this study, we summarize the experience with the use of ultrasound in CH that led us to recommend using an ultrasound-based imaging algorithm [12, 13].

Materials and Methods

Patients

A retrospective review was performed of the cases of all patients whose condition was diagnosed as CH at our institution between January 1, 2003, and December 31, 2010. Only patients whose thyroid ultrasound or ^99mTc-pertechnetate scans were available for review were included. Institutional review board approval was obtained with a waiver of informed consent for the study. All but three CH patients were initially imaged with thyroid ultrasound. The decision to order a ^99mTc-pertechnetate scan was then made by an endocrinologist on the basis of the ultrasound results. Typically ^99mTc-pertechnetate scanning was performed to evaluate or confirm ectopic sublingual thyroid when ultrasound showed thyroid dysgenesis.

TABLE 1:Reference Standard for Thyroid Size (cm) by Age

TABLE 2:Causes of Congenital Hypothyroidism in 124 Patients Between 2003 and 2010

Fig. 1:Photograph shows ideal patient position with hyperextended neck.

Imaging Technique

For thyroid sonography, all patients were examined in the supine position with the neck hyperextended by placement of a folded towel beneath the scapula (Fig. 1). A 7–15 MHz linear transducer with a small footprint was used (Acuson Sequoia 512, Siemens Healthcare, or HDI 5000 IU 22, Philips Healthcare). Gray-scale transverse and longitudinal images were obtained from the base of the tongue. CDUS was performed in some patients to better depict ectopic sublingual thyroid.

For ^99mTc-pertechnetate scintigraphy, the scan was performed with 1–2 mCi of ^99mTc-pertechnetate IV (dose calculated on basis of patient’s weight). Images were obtained in the anterior and lateral views 15 minutes after administration.

Imaging Evaluation and Data Analysis

All of the imaging studies were reviewed at our standard clinical PACS workstation (Synapse, Fujifilm). Both ultrasound and ^99mTc-pertechnetate scans were separately and independently reviewed by a pediatric radiologist (fellowship trained with 5 years of experience) and a nuclear medicine physician (30 years of experience).

Ultrasound studies were reviewed for the presence (eutopic, ectopic, or agenesis) of thyroid tissue, size (normal, hypoplastic, or hyperplastic) compared with the reference standard (Table 1) [14], echotexture (normal or increased echogenicity), and degree of thyroid vascularity (normal, increased, decreased). Technetium-99m-pertechnetate scans were evaluated for the presence (eutopic, ectopic or agenesis) of thyroid tissue and subjective degree (normal, increased, or decreased) of radiotracer uptake.

We used descriptive statistical analysis for each modality, divided into eutopic location, ectopic location, and agenesis of the thyroid gland. We also compared the sensitivity, specificity, and accuracy of the modalities using ^99mTc scintigraphic results (when available) as the reference standard. On ultrasound images we evaluated the presence, location, size, echotexture, and vascularity of thyroid gland, and on the ^99mTc-pertechnetate studies—the reference standard for evaluation of sublingual thyroid— we evaluated presence, location, size, and uptake.

Discussion

The treatment of CH patients is empiric and not guided by imaging findings. A neonate with a diagnosis of CH is immediately treated with thyroid hormone replacement [15]. Using a higher starting dose to more quickly normalize TSH levels to the target range within 2 weeks to normalized developmental IQ even in patients with severe CH is the main purpose of treatment. The initial thyroid hormone (levothyroxine) dose for eutopic thyroid gland is approximately 10 μg/kg/d, compared with 15 μg/kg/d for noneutopic thyroid gland [16].

Permanent CH can be assumed if ultrasound or radionuclide imaging shows the thyroid gland is absent or ectopic (together referred to as dysgenesis) or if at any time during the first year of life, the serum TSH concentration rises above 20 mU/L owing to undertreatment. The American Academy of Pediatrics and the European Society for Pediatric Endocrinology recommend that if permanent CH has not been established by 2–3 years of age, a 30-day trial without thyroid hormone be undertaken. If low serum T4 and elevated TSH concentrations are found, permanent CH is confirmed, and therapy is restarted [1]. If a patient has a eutopic thyroid gland, and the gland produces adequate thyroid hormone in the 30-day trial, a diagnosis of transient CH is established, and the patient needs no further thyroid hormone replacement.

Thyroid scintigraphy is considered the reference standard for the evaluation of CH. There are several reports, however, of limitations of scintigraphy in diagnosing eutopic thyroid gland. Perry et al. [3] and several other groups [5, 17, 18] reported that ultrasound shows thyroid tissue in 2–15% of patients who have none visualized at scintigraphy. These studies included patients with maternal thyrotropin receptor–blocking antibodies, exposure to maternal antithyroid medications, iodine deficiency, or iodine excess causing transient hypothyroidism [1, 5, 15, 17, 18]. These patients may have transient hypothyroidism and would likely not need thyroid hormone replacement for life. Therefore, in patients with a eutopic thyroid gland, no uptake at scintigraphy may be misleading.

In our 7-year cohort of patients with CH, we found a higher frequency of eutopic thyroid gland than reported in the literature. This finding may be related to a higher percentage (36%) of transient hypothyroidism in our screening program [8]. Regardless of the underlying cause, the initial treatment of CH is the same. However, patients with eutopic thyroid gland may need further investigation to differentiate between the permanent and transient forms of CH.

The reported incidence of primary hypothyroidism has increased in the United States over the last two decades. One of the possible causes is inclusion of more cases of transient hypothyroidism. It is important to have a precise diagnosis of eutopic thyroid gland because this can lead to change in treatment [15]. Therefore, some authors advocate the use of both scintigraphy and ultrasound for optimal functional and anatomic detail [5, 8, 19].

Several investigators who used both scintigraphy and ultrasound for the diagnosis of CH have recommended ultrasound as a first-line study to avoid radiation associated with scintigraphy [17, 20, 21]. We are the first, to our knowledge, to report experience using thyroid ultrasound as the primary imaging evaluation of CH with selective use of scintigraphy in children with thyroid dysgenesis.

The main limitation of ultrasound is decreased sensitivity in the evaluation of ectopic thyroid gland. The ultrasound diagnosis of ectopic thyroid gland depends on technique and the experience of the sonographer. Marked variation in sensitivity (0–80%) has been reported among medical centers [1–3, 5, 11, 18, 19]. The sensitivity of sonography in the detection of ectopic thyroid gland in our series was 44%. Using CDUS increases the sensitivity of diagnosis of ectopic thyroid gland [1, 2, 5, 11]. In our series, in most cases of missed ectopic thyroid gland, CDUS was not used. The reported specificity of ultrasound in the detection of ectopic thyroid gland is high [1–3, 5, 11, 18, 19, 21]. In our series, the specificity of sonography was 100%.

Our experience showed that when ultrasound depicts ectopic or eutopic thyroid gland, the scintigraphic results will not change the initial management. For precise diagnosis of agenesis versus sublingual thyroid gland in all patients with ectopic thyroid gland, scintigraphy can be used selectively when ultrasound does not depict any thyroid tissue. In our series, that would have obviated scintigraphy for 54% of the patients.

For management guidance, it is important to differentiate patients with eutopic thyroid gland from those with thyroid dysgenesis. Patients with thyroid dysgenesis are being treated for life with thyroid hormone replacement. The ectopic thyroid gland eventually involutes owing to suppression of TSH. Differentiation between thyroid agenesis and ectopic thyroid gland does not change management. Scintigraphy can therefore be used selectively only in cases of equivocal ultrasound findings, such as hypoplastic thyroid gland. In our series, we did not perform scintigraphy for most patients with eutopic thyroid gland and therefore do not have a correlation with thyroid size or parenchymal echotexture. With this management, we would remove the need for scintigraphy for 90% of patients. This approach will save both radiation and cost with no change in management.

Imaging of patients with CH has a role in the evaluation of the cause, in prognosis, and in guiding management. Ultrasound of the thyroid can be used to differentiate patients with thyroid dysgenesis from patients with eutopic thyroid. Thyroid dysgenesis is typically a sporadic disorder and carries no recurrence risk of CH with future pregnancies. Patients with eutopic thyroid gland are a heterogeneous group; some have a risk of recurrence in future pregnancies. Genetic consultation can be considered [1].

Our study had several limitations. First, our study was performed as a retrospective review of imaging findings, and there was inconsistent use of CDUS, possibly decreasing sensitivity in the detection of ectopic sublingual thyroid gland. Second, the ultrasound and ^99mTc studies were reviewed by a single radiologist, possibly biasing interpretation of the studies. However, compared with original reports, in only two studies (3%) did the retrospective evaluations vary. Because the studies were reviewed by a single radiologist, we could not assess interobserver variability. Third, we imaged only patients who were evaluated prenatally at our institute. Fourth, the study did not include follow-up on euthyroid patients. However, the incidence of transient hypothyroidism in our institution (36%) had been published [8].

Conclusion

Ultrasound can be used as the primary imaging modality for guiding treatment of children with CH, potentially decreasing radiation exposure and cost. Scintigraphy can be reserved for the few patients with equivocal ultrasound findings, such as hypoplastic thyroid gland.

AJR:199, September 2012