Table 6-1^2–4 lists the isotopes most commonly used for in vivo studies of thyroid function. Isotopes with slower physical decay, such as ¹²⁵I and ¹³¹I, are particularly suitable for long-term studies. Conversely, isotopes with faster decay, such as ¹²³I and ¹³²I, usually deliver a lower radiation dose and are advantageous in short-term and repeated studies. Because the peak photon energy γ-emission differs among isotopes, simultaneous studies can be performed with two different isotopes.

Table 6-1

Commonly Used Isotopes for in Vivo Studies and Radiation Dose Delivered

^*Calculations take into account the rate of maximal uptake and the residence time of the isotope, as well as gland size. For the iodine isotopes, average data for adult euthyroid people used were a uptake of 5 hours, a biological of 50 days, maximal uptake of 20%, and gland size of 15 g (see also Quimby et al.⁴ and MIRD^2,3).

^†Dose used for early thyroid uptake studies.

Thyroidal Radioiodide Uptake

The iodide radioisotope usually is given orally in a capsule or in liquid form, and the quantity that is accumulated by the thyroid gland at various intervals is measured with a gamma scintillation counter. It is important to correct for the background activity of isotope circulating in the blood of the neck region (particularly during the early periods after administration). Background correction is achieved by subtracting counts obtained over the region of the thigh. A dose of the same radioisotope, usually 10%, placed in a neck “phantom” is also counted as a “standard.” The percentage of RAIU is calculated from the counts accumulated per constant time unit.

The percentage of RAIU 24 hours after the administration of radioiodide is most useful because in most instances, the thyroid gland has reached the plateau of isotope accumulation, and the best separation between high, normal, and low uptake is obtained at this time. Normal values for 24 hour RAIU in most parts of North America are 5% to 30%. In many other parts of the world, normal values range from 15% to 50%. Lower normal values are due to the increase in dietary iodine intake after the enrichment of foods, particularly mass-produced bread (150 µg of iodine per slice) containing this element. Over the past 3 decades, the mean ingestion of dietary iodine in the United States, although still within the recommended minimum for adults of 125 µg/day, has dramatically declined to approximately 240 to 300 µg/day for men and 190 to 210 µg/day for women.⁵ The inverse relationship between the daily dietary intake of iodine and the RAIU test is clearly illustrated in Fig. 6-1. Therefore, normal values of RAIU uptake will depend on the iodine content in a geographic region and also are related to age (with children having a higher iodine intake than adults). In Japan, the mean dietary iodine is six times higher than in the United States.

FIGURE 6-1 Relationship of 24-hour thyroidal radioiodide (¹³¹I) uptake (RAIU) to dietary content of stable iodine (¹²⁷I). Uptake increases with decreasing dietary iodine. If iodine intake is below the amount provided from thyroid hormone degradation, the latter contributes a larger proportion of the total iodine taken up by the thyroid. With dietary habits in the United States, the average 24 hour thyroidal RAIU is below 20%. (Data from DeGroot LJ, Reed Larsen P, Hennemann G, et al: The Thyroid and Its Diseases. New York, John Wiley & Sons, 1984.)

The intake of large amounts of iodide (>5 mg/day), mainly from the use of iodine-containing radiologic contrast media, antiseptics, vitamins, and drugs such as amiodarone, suppresses RAIU values to a level that is hardly detectable with the usual equipment and doses of isotope. Depending on the type of iodine preparation and the period of exposure, depression of RAIU can last for weeks, months, or even years. Even external application of iodide can suppress RAIU. It therefore is important to inquire about individual dietary habits and sources of excess iodide intake. Because dietary assessment of iodine ingestion can be somewhat inaccurate owing to the variable content of iodine added to various foods, measurement of iodine excretion is a more accurate assessment of the iodine balance. Spot urine iodine measurements were compiled from 1971 to 1974 and from 1988 to 1994 in the National Health and Nutrition Examination Surveys I and III, respectively, and have been found to be decreasing, in accordance with what was stated.⁶ Clinically, if one suspects that the patient had a large iodine load prior to an RAIU, a urine iodine measurement can be obtained. Urine iodine concentrations greater than 100 µg/day usually are associated with RAIU of 20% or less. Therefore, urine iodine can be useful in determining the feasibility of using RAIU.

RAIU is a measure of the avidity of the thyroid gland for iodide and its rate of clearance relative to the kidney, but results of this test do not equate with hormone production or release. Disease states resulting in excessive production of thyroid hormone most often are associated with increased thyroidal RAIU, and those causing hormone underproduction generally are associated with decreased thyroidal RAIU (Fig. 6-2). Some important exceptions to these rules include the high uptake values that are seen in certain hypothyroid patients and the low values noted in some hyperthyroid patients. Increased thyroidal RAIU with hormonal insufficiency can be caused by severe iodide deficiency and by most inborn errors of hormonogenesis. Lack of substrate in the former and specific enzymatic block of hormone synthesis in the latter cause hypothyroidism that is poorly compensated by TSH-induced thyroid gland overactivity. The increase in serum TSH, in response to the low circulating level of thyroid hormone, stimulates thyroidal iodine uptake by the NIS and hence increases RAIU. This can be a point of confusion for the clinician who is confronted with an increased RAIU in a patient who is suspected to have thyroiditis on the basis of blood tests. Alternatively, decreased thyroidal RAIU with hormonal excess typically is encountered in the syndrome of transient thyrotoxicosis (both deQuervain’s and painless thyroiditis) after the ingestion of exogenous hormone (thyrotoxicosis factitia), with iodide-induced thyrotoxicosis (Jod-Basedow disease), rarely in patients with metastatic functioning thyroid carcinoma or struma ovarii, and in patients with thyrotoxicosis who have a moderately high intake of iodide. High or low thyroidal RAIU as a result of low or high dietary iodine intake, respectively, might not be associated with significant changes in thyroid hormone secretion.

FIGURE 6-2 Examples of thyroidal radioiodide uptake curves under various pathologic conditions. Note the prolonged uptake in renal disease caused by decreased urinary excretion of the isotope and the early decline in thyroidal radioiodide content in some patients with thyrotoxicosis associated with a small but rapidly turning over intrathyroidal iodine pool. (Data from DeGroot LJ, Reed Larsen P, Hennemann G, et al: The Thyroid and Its Diseases. New York, John Wiley & Sons, 1984.)

Various factors, including diseases that affect the value of the 24 hour thyroidal RAIU, are listed in Table 6-2. Several variations of the RAIU test have been devised that have particular value under special circumstances. Some of these variations are briefly described.

Table 6-2

Diseases and Other Factors That Affect 24 Hour Thyroidal RAIU

Increased RAIU

Hyperthyroidism (Graves’ disease, Plummer’s disease, toxic adenoma, trophoblastic disease, resistance to thyroid hormone, TSH-producing pituitary adenoma)

Nontoxic goiter (endemic, inherited biosynthetic defects, generalized resistance to thyroid hormone, Hashimoto’s thyroiditis)

Excessive hormonal loss (nephrosis, chronic diarrhea, hypolipidemic resins, diet high in soybean)

Decreased renal clearance of iodine (renal insufficiency, severe heart failure)

Recovery of the suppressed thyroid (withdrawal of thyroid hormone and antithyroid drug administration, subacute thyroiditis, iodine-induced myxedema)

Iodine deficiency (endemic or sporadic dietary deficiency, excessive iodine loss as in pregnancy or in the dehalogenase defect)

TSH administration

Decreased RAIU

Hypothyroidism (primary or secondary)

Syndromes of TSH resistance

Thyroid dysgenesis (hypoplasia, ectopy, or agenesis)

Na/I symporter defect

Defect in iodide concentration (inherited trapping defect, early phase of subacute thyroiditis, transient hyperthyroidism)

Suppressed thyroid gland caused by thyroid hormone (hormone replacement, thyrotoxicosis factitia, struma ovarii)

Iodine excess (dietary, drugs, and other iodine contaminants)

Miscellaneous drugs and chemicals (see Tables 6-10 and 6-13)

RAIU, Radioactive iodine uptake; TSH, thyroid-stimulating hormone.

Early Thyroid Radioiodide Uptake and ^99mTc Uptake Measurements

The combination of severe thyrotoxicosis and a low intrathyroidal iodine concentration may result in an accelerated turnover rate of iodine in some patients. This produces a rapid initial uptake of radioiodide, which reaches a plateau before 6 hours, followed by a decline through release of the isotope in hormonal or other forms (see Fig. 6-2). Although this phenomenon is rare, some laboratories choose to routinely measure early RAIU, usually at 2, 4, or 6 hours. As was mentioned above, early measurements require accurate determination of the background activity contributed by circulating isotope. Radioisotopes with a shorter half-life, such as ¹²³I and ¹³²I, are more suitable in this context.

Because thyroidal uptake in the very early period after administration of radioiodide reflects mainly iodide trapping activity, ^99mTc as the pertechnetate ion (^99mTcO₄⁻) may be used. In euthyroid patients, thyroid trapping is maximal at about 20 minutes and is approximately 1% of the administered dose. This test, when coupled with the administration of triiodothyronine (T₃), has been used to evaluate thyroid gland suppressibility in thyrotoxic patients who are treated with antithyroid drugs.

Perchlorate Discharge Test

The perchlorate discharge test is used to detect defects in intrathyroidal iodide organification. It is based on the following physiologic principle. Iodide is “trapped” in the thyroid gland by an active transport mechanism that is mediated by NIS.⁷ Once in the gland, iodine is rapidly bound to thyroglobulin (Tg), and retention no longer requires active transport. Several ions, such as thiocyanate (SCN⁻) and perchlorate (ClO₄⁻), inhibit NIS-mediated iodide transport and cause release of the intrathyroidal iodide that is not bound to thyroid protein. Thus, intrathyroidal radioiodine loss after the administration of an inhibitor of iodide trapping measures intrathyroidal iodide that is not protein bound and indicates the presence of an iodide-binding defect.

In the standard test, epithyroid counts are obtained every 10 or 15 minutes after the administration of radioiodide. Two hours later, 1 g of KClO₄ is administered orally, and repeated epithyroid counts continue to be obtained for an additional 2 hours. We generally give this dose to patients 6 years or older and give 500 mg to children between the ages of 2 and 6 years. In normal individuals, radioiodide accumulation in the thyroid gland ceases after administration of the iodide transport inhibitor, and little or no loss of the accumulated thyroidal radioactivity occurs after induction of the “trapping” block. An organification defect is indicated if a loss of 5% or more is noted. The severity of the defect is proportional to the extent of radioiodide discharged from the gland and is complete when virtually all the activity accumulated by the gland is lost. The test is positive in the inborn defect of iodide organification caused by thyroid peroxidase (TPO) defects or by mutations in the chloride/iodide transport protein (pendrin) when associated with sensorineural deafness known as Pendred’s syndrome and in defects of the H₂O₂-generating enzyme, dual oxidase 2 (DUOX2), or its maturation factor, DUOXA2. The test may also be positive during the administration of iodide organification–blocking agents or after treatment with radioactive iodide.

Iodine Saliva-to-Plasma Ratio Test

An abnormal iodine (I⁻) saliva-to-plasma (S/P) ratio is pathognomonic of the iodine-trapping defect. The test can be carried out without interruption of thyroid hormone treatment. Furthermore, the measurement of I⁻ S/P can distinguish between a trapping defect and thyroid agenesis, which cannot be determined by RAIU. The I⁻ S/P ratio can be measured in a medical center without access to a gamma camera. The test is based on the observation that all tissues that normally concentrate iodide are affected by the trapping defect.⁸ The presence of an I⁻ transport defect in the parietal cells of the stomach and the choroid plexus of these patients has been used diagnostically by measurement of the gastric fluid-to-plasma and cerebrospinal fluid (CSF)-to-plasma ratios of radioiodide, following the administration of isotope.

One hour after the oral administration of 5 µCi of Na¹²⁵I, saliva is collected without stimulation over a period of 5 to 10 minutes. At the same time, a venous blood sample is obtained. After defrothing of the saliva and removal of cell debris by centrifugation (approximately 10 minutes at 500 × g) and separation of the serum, the S/P ratio of radioiodide is determined by counting equal volumes of these fluids in a gamma scintillation counter. A normal I⁻ S/P is 25. Affected individuals who are not able to concentrate iodide and therefore cannot secrete it into their saliva have a very low S/P. An I⁻ S/P of approximately 1 is diagnostic of a complete trapping defect, and a value between 1 and 20 is consistent with a partial defect.

Measurement of Hormone Concentration and Other Iodinated Compounds and Their Transport in Blood

The tests most commonly used for evaluating thyroid hormone–dependent metabolic status consist of measurements of free thyroid hormone concentrations. This approach is used because of the development of simple, sensitive, and specific methods for measuring these iodothyronines, and because of the lack of specific tests for direct measurement of the metabolic effects of these hormones on target tissues. Other advantages are the requirement of only a small blood sample and the large number of determinations that can be completed by a laboratory during a regular workday. In fact, although a clinician may entertain a diagnosis of thyroid dysfunction, the certainty of the diagnosis is confirmed only after measurement of the thyroid hormone concentrations and thyrotropin (TSH, see later).

The principal source of all hormonal iodine-containing compounds or their precursors is the thyroid gland, whereas peripheral tissues are the source of the products of their degradation. Their chemical structures and normal concentrations in serum are given in Fig. 6-3. It is important to note that the concentration of each substance is dependent not only on the amount synthesized and secreted by the thyroid gland, but also on its affinity for carrier serum proteins, distribution in tissues, rate of degradation, and, finally, clearance.

FIGURE 6-3 Iodine-containing compounds in the serum of healthy adults. a, Iodothyronine concentrations in the euthyroid population are not normally distributed. Therefore, calculation of the normal range on the basis of 95% confidence limits for a Gaussian distribution is accurate. b, Significant decline with old age. c, Probably an overestimation because of cross-reactivity by related substances.

Quantitatively, the major secretory product of the thyroid gland is thyroxine (T₄), with T₃ being next in relative abundance. They are synthesized and stored in the thyroid gland as part of a larger molecule, Tg, which is degraded to release the two iodothyronines in a ratio favoring T₄ by 10- to 20-fold. Under normal circumstances, only minute amounts of Tg escape into the circulation. On a molar basis, it is the least abundant iodine-containing compound in blood. With the exception of T₄, Tg, and small amounts of diiodotyrosine (DIT) and monoiodotyrosine (MIT), all other iodine-containing compounds that are found in normal human serum are produced mainly in extrathyroidal tissues by a stepwise process of deiodination of T₄. An alternative pathway of T₄ metabolism that involves deamination and decarboxylation but retention of the iodine residues gives rise to tetraiodothyroacetic acid (TETRAC) and triiodothyroacetic acid (TRIAC).9,10 Conjugation to form sulfated iodoproteins also occurs. Sulfoconjugates of T₄, T₃, and reverse T₃ (rT₃) have been identified in human biological fluids. Additionally, maternal serum levels of 3,3′-diiothyronine sulfate (T₂S) may reflect on the status of fetal thyroid function. Circulating iodalbumin is generated by intrathyroidal iodination of serum albumin. Small amounts of iodoproteins may be formed in peripheral tissues or in serum by covalent linkage of T₄ and T₃ to soluble proteins. The physiologic function of circulating iodine compounds other than T₄ and T₃ remains unknown, with the exception of rT₃. rT₃ levels are elevated during fasting and during significant nonthyroidal illness. In such instances, measurement of rT₃ can help the clinician to distinguish between these conditions and central hypothyroidism.

Measurement of Total Thyroid Hormone Concentration in Serum

Iodometry

Because iodine is an integral part of the thyroid hormone molecule, it is not surprising that determination of the iodine content in serum was the first method used over 6 decades ago for the identification and quantitation of thyroid hormone.¹¹ Measurement of protein-bound iodine was the earliest method used routinely for the estimation of thyroid hormone concentration in serum. This test measured the total quantity of iodine precipitable with serum proteins, 90% of which is T₄. The normal range was 4 to 8 mg of iodine per deciliter of serum.

Efforts to measure serum thyroid hormone levels with greater specificity and with lesser interference from nonhormonal iodinated compounds led to the development of measurement of butanol-extractable iodine and T₄ iodine by column techniques. All such chemical methods for the measurement of thyroid hormone in serum have been replaced by ligand assays, which are devoid of interference by even large quantities of nonhormonal iodine-containing substances.

Table 6-3

Conditions Associated With Changes in Serum TT₄ Concentration and Relationship to Clinical Status

T₃, Triiodothyronine; T₄, thyroxine; TBG, thyroxine-binding globulin; TSH, thyroid-stimulating hormone; TT₄, total T₄.

Serum TT₄ levels are low in conditions that are associated with decreased TBG concentrations, in the presence of abnormal TBGs with reduced binding affinity, and when the available T₄-binding sites on TBG are partially saturated by competing drugs present in blood in high concentration (Table 6-4). Conversely, TT₄ levels are high when the serum TBG concentration is high. In this situation, the person remains euthyroid provided that feedback regulation of the thyroid gland is intact.

Table 6-4

Compounds That Affect Thyroid Hormone Serum Transport Proteins

NSAID, Nonsteroidal antiinflammatory drug; o,p′-DDD, 2,4′-dichlorodiphenyldichloroethane (mitotane); TBG, thyroxine-binding globulin; TTR, transthyretin.

Although changes in transthyretin (TTR) concentration rarely give rise to significant alterations in TT₄ concentration, the presence of a variant serum albumin with high affinity for T₄ or antibodies against T₄ produce apparent elevations in the measured TT₄ concentration, whereas the metabolic status remains normal. The variant albumin is inherited as an autosomal dominant trait termed familial dysalbuminemic hyperthyroxinemia (FDH).

Another possible cause of discrepancy between the observed serum TT₄ concentration and the metabolic status of the patient is divergent changes in serum total T₃ (TT₃) and TT₄ concentrations with alterations in the serum T₃/T₄ ratio. The most common situation is that of an elevated TT₃ concentration. The source of T₃ may be endogenous, as in T₃ thyrotoxicosis, or exogenous, as during ingestion of T₃. In the former situation, contrary to the common variety of thyrotoxicosis, elevation in the serum TT₃ concentration is not accompanied by an increase in the TT₄ level. In fact, the serum TT₄ level is normal and occasionally low. This finding indicates that the pathogenesis of T₃ thyrotoxicosis is the direct secretion of T₃ from the thyroid gland rather than the peripheral conversion of T₄ to T₃. Ingestion of pharmacologic doses of T₃ results in thyrotoxicosis associated with severe depression of the serum TT₄ concentration. Moderate hypersecretion of T₃ can be associated with euthyroidism and a low serum TT₄ concentration. This situation, occasionally referred to as T₃ euthyroidism, may be more prevalent than T₃ thyrotoxicosis. It is believed to constitute a state of compensatory T₃ secretion as a physiologic adaptation of the failing thyroid gland, such as after treatment of thyrotoxicosis, in some cases of chronic thyroiditis, or during iodine deprivation. The serum TT₄ concentration is also low in normal persons receiving replacement doses of T₃. Conversely, serum TT₄ levels are above the upper limit of normal in 15% to 50% patients treated with exogenous T₄ and having normal serum TSH. Because of the relatively slow rate of metabolism and the large extrathyroidal T₄ pool, the serum concentration of the hormone varies little with the time of sampling in relation to ingestion of the daily dose.

Serum Total T₃

Triiodothyronine is principally responsible for the effects of thyroid hormones on the target organs. T₃ is formed extrathyroidally via 5′ deiodination of T₄. Thus, serum T₃ concentration reflects the functional state of the peripheral tissue rather than the secretory performance of the thyroid gland. Like T₄, 99% of T₃ is present in protein-bound form. However, the affinity of T₃ to TBG is 10-fold lower than that of T₄.

Normal serum TT₃ concentrations in the adult range from 80 to 190 ng/dL (1.2 to 2.9 nmol/L). Sex differences are small, but age differences are more dramatic. In contrast to serum TT₄, the TT₃ concentration at birth is low, about half the normal adult level. It rises rapidly within 24 hours to about double the normal adult value, followed by a decrease over the subsequent 24 hours to a level in the upper adult range, which persists for the first year of life. A decline in the mean TT₃ level has been observed in old age, although not in healthy subjects,23,24 which suggests that a fall in TT₃ might reflect the prevalence of nonthyroidal illness rather than an effect of age alone. Although a positive correlation between serum TT₃ level and body weight has been observed, this might be related to overeating.²⁵ Rapid and profound reductions in serum TT₃ can be produced within 24 to 48 hours of total calorie or carbohydrate-only deprivation.

Most conditions that cause serum TT₄ levels to increase are associated with high TT₃ concentrations. Thus, serum TT₃ levels are usually elevated in thyrotoxicosis and reduced in hypothyroidism. However, in both conditions, the TT₃/TT₄ ratio is elevated relative to normal euthyroid persons. This elevation is due to the disproportionate increase in serum TT₃ concentration in thyrotoxicosis and a lesser diminution in hypothyroidism relative to the TT₄ concentration. Accordingly, measurement of the serum TT₃ level is a more sensitive test for the diagnosis of hyperthyroidism, and measurement of TT₄ is more useful in the diagnosis of hypothyroidism.

Under certain conditions, changes in the serum TT₃ and TT₄ concentrations are disproportionate or occur in the opposite direction (Table 6-5). Such conditions include the syndrome of thyrotoxicosis with normal TT₄ and FT₄ levels (T₃ thyrotoxicosis). In some patients, treatment of thyrotoxicosis with antithyroid drugs normalizes the serum TT₄ but not the TT₃ level and produces a high TT₃/TT₄ ratio. In areas of limited iodine supply and in patients with limited thyroidal ability to process iodide, euthyroidism can be maintained at low serum TT₄ and FT₄ levels by increased direct thyroidal secretion of T₃. Although these changes have a rational physiologic explanation, the significance of discordant serum TT₄ and TT₃ levels under other circumstances is less well understood.

Table 6-5

Conditions That May Be Associated With Discrepancies Between the Concentration of Serum TT₃ and TT₄

TT₃, Total triiodothyronine; TT₄, total thyroxine.

^*Artifactual values depend on the method of hormone determination in serum.

^†Hepatic and renal failure, diabetic ketoacidosis, myocardial infarction, infectious and febrile illness, cancers.

^‡Glucocorticoids, iodinated contrast agents, amiodarone, propranolol, propylthiouracil.

The most common cause of discordant serum concentrations of TT₃ and TT₄ is a selective decrease in serum TT₃ caused by decreased conversion of T₄ to T₃ in peripheral tissues. This reduction is an integral part of the pathophysiology of a number of nonthyroidal acute and chronic illnesses and calorie deprivation. In these conditions, the serum TT₃ level is often lower than that commonly found in patients with frank primary hypothyroidism. However, no clear clinical evidence of hypometabolism is found in this situation. In some individuals, decreased T₄-to-T₃ conversion is an inherited condition.²⁶ A combination of high TT₃ and low TT₄ is typical in subjects with loss-of-function mutations in the iodothyronine cell membrane transporter, MCT8.²⁷

A variety of drugs are responsible for producing changes in the serum TT₃ concentration without apparent metabolic consequences. Drugs that compete with hormone binding to serum proteins decrease serum TT₃ levels, generally without affecting the free T₃ (FT₃) concentration (see Table 6-4). Some drugs such as glucocorticoids²⁸ depress the serum TT₃ concentration by interfering with the peripheral conversion of T₄ to T₃. Others, such as phenobarbital,²⁹ depress the serum TT₃ concentration by stimulating the rate of intracellular hormone degradation and clearance. Most have multiple effects. These effects are combinations of those described above, as well as inhibition of the hypothalamic-pituitary axis or thyroidal hormonogenesis.

Changes in serum TBG concentration have an effect on the serum TT₃ concentration similar to that on TT₄. The presence of endogenous antibodies to T₃ can also result in apparent elevation of serum TT₃, but as in the case of high TBG, it does not cause hypermetabolism.

Administration of commonly used replacement doses of T₃, usually on the order of 75 µg/day or 1 µg/kg body weight per day,³⁰ results in serum TT₃ levels in the thyrotoxic range. Furthermore, because of rapid gastrointestinal absorption and a relatively fast degradation rate, the serum level varies considerably according to the time of sampling in relation to hormone ingestion.

Measurement of Total and Unsaturated Thyroid Hormone–Binding Capacity in Serum

The concentration of thyroid hormone in serum is dependent on its supply, as well as on the abundance of hormone-binding sites on serum proteins; therefore, estimation of the latter has proved useful in the correct interpretation of values obtained from measurement of the total hormone concentration. These results have been used to provide an estimate of the free hormone concentration, which is important in differentiating changes in serum total hormone concentrations caused by alterations in binding proteins in euthyroid patients from those caused by abnormalities in thyroid gland activity that give rise to hypermetabolism or hypometabolism.

In Vitro Uptake Tests

In vitro uptake tests measure the unoccupied thyroid hormone–binding sites on TBG. They use labeled T₃ or T₄ and some form of synthetic absorbent to measure the proportion of radiolabeled hormone that is not tightly bound to serum proteins. Because ion exchange resins often are used as absorbents, the test became known as the resin T₃ or T₄ uptake test, which describes the technique rather than the entity measured.

The test is usually carried out by incubating a sample of the patient’s serum with a trace amount of labeled T₃ or T₄. The labeled hormone, not bound to available binding sites on TBG present in the serum sample, is absorbed onto an anion exchange resin and measured as resin-bound radioactivity. Values correlate inversely with the concentration of unsaturated TBG. Various methods use different absorbing materials to remove the hormone that is not tightly bound to TBG. Labeled T₃ generally is used because of its less firm, yet preferential, binding to TBG. Depending on the method, typical normal results for T₃ uptake are 25% to 35% or 45% to 55%. Thus, it is more valuable to express results of the uptake tests as a ratio of the result obtained in a normal control serum run in the same assay as the test samples. Normal values will then range on either side of 1.0, usually from 0.85 to 1.15.

Uptake of tracer by the absorbent is inversely proportional to the number of unsaturated binding sites (unoccupied by endogenous thyroid hormone) in serum TBG. Thus, uptake is increased when the amount of unsaturated TBG is reduced as a result of excess endogenous thyroid hormone or a decrease in the concentration of TBG. In contrast, uptake is decreased when the amount of unsaturated TBG is increased as a result of a low serum thyroid hormone concentration or an increase in the concentration of TBG. Because the test can be affected by either or both independent variables—serum total thyroid hormone and TBG concentrations—the results cannot be interpreted without knowledge of the hormone concentration. As a rule, parallel increases or decreases in serum TT₄ concentration and the T₃ uptake test indicate hyperthyroidism and hypothyroidism, respectively, whereas discrepant changes in serum TT₄ and T₃ uptake suggest abnormalities in TBG binding. However, abnormalities in hormone and TBG concentrations can coexist in the same patient. For example, a hypothyroid patient with a low TBG level will typically show a low TT₄ level and normal T₃ uptake results (Fig. 6-4). Several nonhormonal compounds, because of structural similarities, compete with thyroid hormone for its binding site on TBG. Some are used as pharmacologic agents and thus may alter the in vitro uptake test, as well as the total thyroid hormone concentration in serum. A list is provided in Table 6-4.

FIGURE 6-4 Graphic representation of the relationship between the serum total thyroxine (T₄) concentration, the resin triiodothyronine uptake (rT₃U) test, and the free T₄ (FT₄) concentration in various metabolic states and in association with changes in thyroxine-binding globulin (TBG). The principle of communicating vessels is used as an illustration. The height of fluid in the small vessel represents the level of FT₄; the total amount of fluid in the large vessel, the total T₄ concentration; and the total volume of the large vessel, the TBG capacity. Dots represent resin beads; black dots represent those carrying the radioactive T₃ tracer (T₃*). The rT₃U test result (black dots) is inversely proportional to the unoccupied TBG-binding sites represented by the unfilled capacity of the large vessel.

TBG and TTR Measurements

The concentrations of TBG and TTR in serum can be estimated by measurement of their total T₄-binding capacity at saturation or measured directly by immunologic techniques.31,32

The TBG concentration in serum can be determined by RIA,³² and both TBG and TTR can be measured by Laurell’s rocket immunoelectrophoresis, by radial immunodiffusion, or by enzyme immunoassay; commercial methods are available. The true mean value for TBG is 1.6 mg/dL (260 nmol/L), with a range of 1.1 to 2.2 mg/dL (180 to 350 nmol/L) in serum. In adults, the normal range for TTR is 16 to 30 mg/dL (2.7 to 5.0 mmol/L). Concentrations of TBG and TTR in serum vary with age, gender, pregnancy, and posture. Determination of the concentration of these proteins in serum is particularly helpful for evaluation of extreme deviations from normal, as in congenital abnormalities of TBG. In most instances, however, the in vitro uptake test, in conjunction with the serum TT₄ level, gives an approximate estimation of the TBG concentration.

Estimation of Free Thyroid Hormone Concentration

Most thyroid hormones in the blood are bound to serum protein carriers, thus leaving only a minute fraction of free hormone in the circulation that is capable of mediating biological activities. A reversible equilibrium exists between bound and unbound hormone, and it is the latter that represents the fraction of the hormone capable of traversing cellular membranes to exert its effects on body tissues. Although changes in serum hormone-binding proteins affect both the total hormone concentration and the corresponding circulating free fraction, in a euthyroid person, the absolute concentration of free hormone remains constant and correlates with the tissue hormone level and its biological effect. Information concerning this value is probably the most important parameter in the evaluation of thyroid function because it relates to the patient’s metabolic status, although other mechanisms exist for the cell to control the active amount of thyroid hormone via autoregulation of receptors ³³ and regulation of deiodinase activity.^34,35 Rarely, a defect in thyroid hormone transport into cells would abolish the free hormone and the metabolic effect correlation.²⁷

With few exceptions, the free hormone concentration is high in thyrotoxicosis, low in hypothyroidism, and normal in euthyroidism, even in the presence of profound changes in TBG concentration, provided that the patient is in a steady state. Notably, the FT₄ concentration may be normal or even low in patients with T₃ thyrotoxicosis and in those ingesting pharmacologic doses of T₃. The concentration of FT₄ may be outside the normal range in the absence of an apparent abnormality in thyroid hormone–dependent metabolic status. This situation is frequently observed in severe nonthyroidal illness, during which both high and low values have been reported. As expected, when a euthyroid state is maintained by the administration of T₃ or by predominant thyroidal secretion of T₃, the FT₄ level is also depressed. More consistently, patients with a variety of nonthyroidal illnesses have low FT₃ levels. This decrease is characteristic of all conditions associated with depressed serum TT₃ concentrations caused by diminished conversion of T₄ to T₃ in peripheral tissues by deiodinase enzymes. Both FT₄ and FT₃ values may be out of line in patients receiving a variety of drugs (see below). Marked elevations in both FT₄ and FT₃ concentrations in the absence of hypermetabolism are typical of patients with the inherited condition of resistance to thyroid hormone. The FT₃ concentration is usually normal or even high in hypothyroid individuals living in areas of severe endemic iodine deficiency. Their FT₄ levels are, however, normal or low. Free hormone concentrations also do not reflect the metabolic status of the patient with inherited defects in hormone transport into cells of hormone metabolism.³⁶

Direct Measurement of Free T₄ and Free T₃

Direct measurement of absolute FT₄ and FT₃ concentrations is technically difficult and until recently has been limited to research assays. To minimize perturbations of the relationship between free and bound hormone, these hormones must be separated by ultrafiltration or by dialysis involving minimal dilution and little alteration in pH or electrolyte composition. The separated free hormone is then measured directly by RIA or chromatography.³⁷ These assays are probably the most accurate available, but small, weakly bound, dialyzable substances or drugs may be removed from the binding proteins, and the free hormone concentration measured in their presence might not fully reflect the free concentration in vivo. Direct immunometric assays adapted to automation, although not reliable under specific conditions, have replaced more labor intensive methods (see below).

Isotopic Equilibrium Dialysis

This method has been the gold standard for the estimation of FT₄ or FT₃ for more than 40 years. It is based on a determination of the proportion of T₄ or T₃ that is unbound, or free, and thus is able to diffuse through a dialysis membrane (i.e., the dialyzable fraction). To carry out the test, a sample of serum is incubated with a trace amount of labeled T₄ or T₃. The labeled tracer rapidly equilibrates with the respective bound and free endogenous hormones. The sample is then dialyzed against buffer at a constant temperature until the concentration of free hormone on either side of the dialysis membrane has reached equilibrium. The dialyzable fraction is calculated from the proportion of labeled hormone in the dialysate. The contribution from radioiodide present as contaminant in the labeled tracer hormone should be eliminated by purification ³⁸ and by various techniques of precipitation of the dialyzed hormone.³⁹ FT₄ and FT₃ levels can be measured simultaneously by addition to the sample of T₄ and T₃ labeled with two different radioiodine isotopes. Ultrafiltration is a modification of the dialysis technique. Results are expressed as the fraction (dialyzable fraction of T₄ or T₃) or percentage (%FT₄ or %FT₃) of the respective hormones that dialyzed, and the absolute concentrations of FT₄ and FT₃ are calculated from the product of the total concentration of the hormone in serum and its respective dialyzable fraction. Typical normal values for FT₄ in adults range from 1.0 to 3.0 ng/dL (13 to 39 pmol/L), and those for FT₃ range from 0.25 to 0.65 ng/dL (3.8 to 10 nmol/L).

Results achieved by these techniques generally are comparable to those determined by direct one-step methods (see below) but are more likely to differ with extremely low or extremely high TBG concentrations or in the presence of circulating inhibitors of protein binding, especially in situations of nonthyroidal illness. The measured dialyzable fraction may be altered by the temperature at which the assay is run, the degree of dilution, the time allowed for equilibrium to be reached, and the composition of the diluting fluid. The calculated value is dependent on an accurate measurement of TT₄ or TT₃ and may be incorrect in patients with T₄ or T₃ autoantibodies. Some of these problems, particularly those arising from dilution, can be surmounted by using commercially available dialysis methods or ultrafiltration methods of free from bound hormone that do not necessitate serum dilution. In addition, the antibody that is used on the ultrafiltrate must be of high affinity to achieve acceptable precision.

Index Methods

Because determination of free hormone by equilibrium dialysis is cumbersome and technically demanding, many clinical laboratories have used a method by which an FT₄ index (FT₄I) or an FT₃ index (FT₃I) is derived from the product of the TT₄ or TT₃ (determined by immunoassay) and the value of an in vitro uptake test (see above). Although the results are not always in agreement with the values obtained by dialysis, these techniques are rapid and simple. They are more likely to fail at extremely low or extremely high TBG concentrations, in the presence of abnormal binding proteins, in patients with nonthyroidal illness, or in the presence of circulating inhibitors of protein binding.

The theoretical contention that the FT₄I is an accurate estimate of the absolute FT₄ concentration can be confirmed by the linear correlation between these two parameters. This statement is true provided that results of the in vitro uptake test (T₃ or T₄ uptake) are expressed as the thyroid hormone–binding ratio, which is determined by dividing the tracer counts bound to the solid matrix by counts bound to serum proteins. Values are corrected for assay variations by using appropriate serum standards and are expressed as the ratio of a normal reference pool. The normal range is slightly narrower than the corresponding TT₄ in healthy euthyroid patients with a normal TBG concentration. It is 6.0 to 10.5 mg/dL (77 to 135 nmol/L) when calculated from TT₄ values. The FT₄I is high in thyrotoxicosis and low in hypothyroidism, irrespective of the TBG concentration. Euthyroid patients with TT₄ values outside the normal range as a result of TBG abnormalities have a normal FT₄I. Lack of correlation between the FT₄I and the metabolic status of the patient has been observed in the same circumstances as those described for similar discrepancies when the FT₄ concentration was measured by dialysis.

Methods for estimation of the FT₃I are also available but are rarely used in routine clinical evaluation of thyroid function. Like the FT₄I, it correlates well with the absolute FT₃ concentration. The test corrects for changes in TT₃ concentration resulting from variations in TBG concentration.

Estimation of Free T₄ and Free T₃ Based on TBG Measurements

Because most T₄ and T₃ in serum are bound to TBG, their free concentrations can be calculated from their binding affinity constants to TBG and molar concentrations of hormones and TBG. A simpler calculation of the T₄/TBG and T₃/TBG ratios yields values that are similar to but less accurate than the FT₄I and FT₃I, respectively.

Two-Step Immunoassays: In these assays, the free hormone is first immunoextracted by a specific bound antibody (first step), frequently fixed to the tube (coated tube).⁴⁰ After washing, labeled tracer is added and is allowed to equilibrate between the unoccupied sites on the antibody and those of serum thyroid hormone–binding proteins. The free hormone concentration will be inversely related to the antibody-bound tracer, and values are determined by comparison to a standard curve. Values that are obtained with this technique are generally comparable to those determined by direct methods. They are more likely to differ in the presence of circulating inhibitors of protein binding and in sera from patients with nonthyroidal illness.

Analogue (One-Step) Immunoassays: In these assays, a labeled analogue of T₄ or T₃ directly competes with the endogenous free hormone for binding to antibodies.⁴¹ In theory, these analogues are not bound by the thyroid hormone–binding proteins in serum. However, various studies have found significant protein binding to the variant albumin-like protein, to TTR, and to iodothyronine autoantibodies. Such binding results in discrepant values in other assays in a number of conditions, including nonthyroidal illness, pregnancy, and familial dysalbuminemic hyperthyroxinemia (FDH).⁴² A growing number of commercial kits are available, some of which have been modified to minimize these problems.^43,44 Nonetheless, their accuracy remains controversial, although such commercial methods are increasingly being adopted in the routine clinical chemistry laboratory. Commercially available kits for measurement of free T₄ values are compared in Table 6-6.

Table 6-6

Commercial Free T₄ Methods

Name	Methodology	Manufacturer
Amberlite MAB	Serum free T₄ inhibits binding of peroxidase anti-T₄ monoclonal antibody T₃-coated solid phase.	Amersham, UK
Chiron ACS:180	Serum free T₄ competes with acridinium ester labeled T₄. Anti-T₄ antibody linked to magnetic particles.	Chiron Diagnostics, MA, USA
AxSYM	Anti-T₄ coated microparticles. T₃-alkaline phosphatase binds to the unoccupied sites.	Abbott Labs, IL, USA
Elecsys	Anti-T₄ antibody labeled with ruthenium. Unoccupied antibody binds to biotinylated T₄, which is linked to streptavidin-coated microparticles. Magnetic separation.	Boehringer Manheim, IN, USA
Diagnostic Product Immulite	T₄ analogue tracer (does not bind TBG or TTR) competes with serum free T₄ for a limited number of T₄ antibody binding sites. Alkaline phosphatase–labeled antianalogue binds to solid phase, and generated signal is inversely proportional to free T₄.	Diag Prod, CA, USA
Corning Nichols Dialysis	Dialysis against 12-fold buffer volume, followed by radioimmunoassay of T₄ in the dialysate.	Nichols Institute, CA, USA

Automated Measurement of Free T₄ and Free T₃: During the 1990s, through the introduction of random access immunoassay analyzers that operate with chemiluminescent or fluorescent labels, measurements of free thyroid hormones became automated and therefore allowed rapid processing of multiple samples. Although the initial financial burden of such equipment is considerable, they reduce labor costs, demand few handling skills on behalf of the operator, and provide random access so that samples can be tested on demand. Precision studies have shown highly reproducible data with this approach.45,46 Comparison of results between different automated analyzers and with manual free thyroid hormone assays, including the gold standard of equilibrium dialysis, has revealed good correlation over a broad range of free thyroid hormone concentrations.^47,48

Considerations in Selection of Methods for the Estimation of Free Thyroid Hormone Concentration

No single method for the estimation of free hormone concentration in serum is infallible in the evaluation of thyroid hormone–dependent metabolic status. Each test has inherent advantages and disadvantages depending on specific physiologic and pathologic circumstances. For example, methods based on measurement of total thyroid hormone and TBG cannot be used in patients with absent TBG secondary to inherited TBG deficiency. Under such circumstances, the concentration of free thyroid hormone is dependent on interaction of the hormone with serum proteins that normally play a negligible role (TTR and albumin). When alterations in thyroid hormone binding do not affect T₄ and T₃ equally, discrepant results of FT₄I are obtained when labeled T₄ or T₃ is used in the in vitro uptake test. For example, euthyroid patients with FDH and those who have endogenous antibodies with greater affinity for T₄ will have high TT₄ but a normal T₃ uptake test, which will result in an overestimation of the calculated FT₄I. In such instances, calculation of the FT₄I from a T₄ uptake test may provide more accurate results. Conversely, reduced overall binding affinity for T₄, which affects T₃ to a lesser extent, will underestimate the FT₄I derived from a T₃ uptake test. Similarly, use of T₄ and T₃ uptake for estimation of the free hormone concentration is satisfactory in the presence of alterations in TBG concentration but not with alterations of the affinity of TBG for the hormone.

Methods based on equilibrium dialysis are most appropriate for estimation of the free thyroid hormone level in patients with all varieties of abnormal binding to serum proteins, provided that the true concentration of total hormone has been accurately determined. All methods for the estimation of FT₄ may give high or low values in patients with severe nonthyroidal illness who are believed to be euthyroid. This finding has been attributed, at least in part, to the presence of inhibitors of thyroid hormone binding to serum proteins, as well as to the various adsorbents that are used in the test procedures. Some of these inhibitors have been postulated to leak from tissues of the diseased patient. Such discrepancies are even more pronounced during transient states of hyperthyroxinemia or hypothyroxinemia associated with acute illness, after withdrawal of treatment with thyroid hormone, and in patients with acute changes in TBG concentration.

The contribution of various drugs that interfere with binding of thyroid hormone to serum proteins or with the in vitro tests should also be taken into account in the choice and interpretation of tests (see Table 6-4). Although the free thyroid hormone concentration in serum would appear to determine the amount of hormone that is available to body tissues, factors that govern their cell membrane uptake, transport to the nucleus, and functional interactions with nuclear receptors and cofactors ultimately determine their biological effects.

Measurements of Iodine-Containing Hormone Precursors and Products of Degradation

The last 4 decades have witnessed the development of RIAs for the measurement of a number of naturally occurring, iodine-containing substances that have little, if any, thyromimetic activity. Some of these substances are products of T₄ and T₃ degradation in peripheral tissues. Others are predominantly, if not exclusively, of thyroidal origin. Because they are devoid of significant metabolic activity, with the exception of rT₃, measurement of their concentration is of value only in the research setting for detecting abnormalities in the metabolism of thyroid hormone in peripheral tissues, as well as defects of hormone synthesis and secretion. The application of tandem mass spectrometry in the measurement of these substances undoubtedly will result in a better understanding of their changes and their pathophysiologic implications.

3,3′,5′-Triiodothyronine or Reverse T₃

rT₃ is principally a product of T₄ degradation in peripheral tissues, namely, liver and kidney. It is also secreted by the thyroid gland, but the amounts are practically insignificant.⁴⁹ rT₃ is an inactive product of T₄ degradation. Thus, measurement of the rT₃ concentration in serum reflects both tissue supply and metabolism of T₄ and identifies conditions that favor this particular pathway of T₄ degradation.

When total rT₃ (TrT₃) is measured in unextracted serum, a competitor of rT₃ binding to serum proteins must be added. Several chemically related compounds may cross-react with the antibodies. The strongest cross-reactivity is observed with 3,3′-diiodothyronine (3,3′-T₂), but such cross-reactivity does not present a serious methodologic problem because of its relatively low level in human serum. Although cross-reactivity with T₃ and T₄ is less likely, these compounds are more often the cause of rT₃ overestimation because of their relative abundance, particularly in thyrotoxicosis. Free fatty acids interfere with the measurement of rT₃ by RIA.⁵⁰ The normal range in adult serum for TrT₃ is 14 to 30 ng/dL (0.22 to 0.46 nmol/L), although varying values have been reported. It is elevated in subjects with high TBG and in some individuals with FDH.⁵¹ Serum TrT₃ levels are normal in hypothyroid patients who are treated with T₄, which indicates that peripheral T₄ metabolism is an important source of circulating rT₃. Values are high in thyrotoxicosis and low in untreated hypothyroidism. High values are normally found in cord blood and in newborns.

With only a few exceptions, notably uremia and human immunodeficiency virus (HIV) infection and acquired immunodeficiency syndrome (AIDS), serum TrT₃ concentrations are elevated in all circumstances that cause low serum T₃ levels in the absence of obvious clinical signs of hypothyroidism. These conditions include, in addition to the newborn period, a variety of acute and chronic nonthyroidal illnesses, calorie deprivation, and the influence of a growing list of clinical agents and drugs (Table 6-7).

Table 6-7

Agents That Alter the Extrathyroidal Metabolism of Thyroid Hormone

PTU, Propylthiouracil; T₃, triiodothyronine; T₄, thyroxine.

The current clinical application of TrT₃ measurement in serum is in the differential diagnosis of conditions associated with alterations in serum T₃ and T₄ concentrations when thyroid gland and metabolic abnormalities are not readily apparent.

The dialyzable fraction of rT₃ in normal adult serum is 0.2% to 0.32%, or approximately the same as that of T₃. The corresponding serum free rT₃ (FrT₃) concentration is 50 to 100 pg/dL (0.77 to 1.5 pmol/L). In the absence of gross TBG abnormalities, variations in serum FrT₃ concentration closely follow those of TrT₃. Reverse T₃ is measured by RIA in serum, plasma, or amniotic fluid samples. Antigen from patient samples competes with radioactive tracer (¹²⁵I-rT₃) for binding sites on an antibody. After incubation, the amount of tracer that is bound to antibody will be inversely proportional to the amount of antigen in the sample. The radioactivity of the tracer bound to antibody is counted, and the amount of antigen or rT₃ from the patient’s serum is determined from this. Typically, 0.1 mL of undiluted sample is required.

3,5-Diiodothyronine

The normal adult range for total 3,5-diiodothyronine (3,5-T₂) in serum measured by direct RIAs is 0.20 to 0.75 ng/dL (3.8 to 14 pmol/L). That 3,5-T₂ is derived from T₃ is supported by the observations that conditions associated with high and low serum T₃ levels have elevated and reduced serum concentrations of 3,5-T₂, respectively. Thus, high serum 3,5-T₂ levels have been reported in hyperthyroidism, and low levels have been reported in the serum of hypothyroid patients, in newborns, during fasting, and in patients with liver cirrhosis.

3,3′-Diiodothyronine

Normal concentrations in adults probably range from 1 to 8 ng/dL (19 to 150 pmol/L). Levels are clearly elevated in hyperthyroidism and in the newborn. Values have been found to be normal or depressed in nonthyroidal illnesses, in agreement with the demonstration of reduced monodeiodination of rT₃ to 3,3′-T₂. In vivo turnover kinetic studies and measurement of 3,3′-T₂ in serum after the administration of T₃ and rT₃ have clearly shown that 3,3′-T₂ is the principal metabolic product of these two triiodothyronines.

3′,5′-Diiodothyronine

Reported concentrations of 3′,5′-diiodothyronine (3′,5′-T₂) in the serum of normal adults have a mean overall range of 1.5 to 9.0 ng/dL (30 to 170 pmol/L).52,53 Values are high in hyperthyroidism and in the newborn. Being the derivative of rT₃ monodeiodination, 3′,5′-T₂ levels are elevated in serum during fasting and in chronic illness, in which the level of the rT₃ precursor is also high. Administration of dexamethasone also produces an increase in the serum 3′,5′-T₂ level.

3′-Monoiodothyronine

The concentration of 3′-monoiodothyronine (3′-T₁) in the serum of normal adults, as measured by RIA, has been reported to range from 0.6 to 2.3 ng/dL (15 to 58 pmol/L) and from less than 0.9 to 6.8 ng/dL (<20 to 170 pmol/L). Its two immediate precursors, 3,3′-T₂ and 3′,5′-T₂, are the main cross-reactants in the RIA. Serum levels are very high in hyperthyroidism and low in hypothyroidism. The concentration of 3′-T₁ in serum is elevated in all conditions associated with high rT₃ levels, including the newborn period, nonthyroidal illness, and fasting. This finding is not surprising because the immediate precursor of 3′-T₁ is 3′,5′-T₂, a product of rT₃ deiodination, which is also present in serum in high concentration under the same circumstances. Elevated serum levels of 3′-T₁ in renal failure are attributed to decreased clearance inasmuch as the concentrations of its precursors are not increased.

3-Monoiodothyronine

Experience with the measurement of 3-T₁ in serum is limited. Normal values in the serum of adult humans determined by ³H-labeled 3-T₁ in a specific RIA ranged from less than 0.5 to 7.5 ng/dL (<13 to 190 pmol/L). The mean concentration of 3-T₁ in the serum of thyrotoxic patients and in cord blood was significantly higher; 3-T₁ appears to be a product of in vivo deiodination of 3,3′-T₂.

Tetraiodothyroacetic Acid and Triiodothyroacetic Acid

The iodoamino acids TETRAC (T₄A) and TRIAC (T₃A), products of deamination and oxidative decarboxylation of T₄ and T₃, respectively, have been detected in serum by direct RIA measurements. Reported mean concentrations in the serum of healthy adults have been 8.7 ng/dL ⁵⁴ and 2.6 ng/dL (range, 1.6 to 3.0 ng/dL or 26 to 48 pmol/L)²⁵ for T₃A, and 28 ng/dL (range, <8 to 60 mg/dL or <105 to 800 pmol/L) for T₄A. Serum T₄A levels are reduced during fasting and in patients with severe illness, although the percentage of conversion of T₄ to T₄A is increased.^55,56 The concentration of serum T₃A remains unchanged during the administration of replacement doses of T₄ and T₃.²⁵ It has been suggested that intracellular rerouting of T₃ to T₃A during fasting is responsible for the maintenance of normal serum TSH levels in the presence of low T₃ concentrations.

3,5,3′-Triiodothyronine Sulfate and 3,3′-Diiodothyronine Sulfate

Sulfation of iodothyronines results in the inactivation of thyroid hormones and enhances their excretion in urine and bile. An RIA procedure is available to measure 3,5,3′-triiodothyronine sulfate (T₃S) in ethanol extracted serum samples. Concentrations of T₃S in normal adults range from 4 to 10 ng/dL (50 to 125 pmol/L). Although the principal source of T₃S is T₃ and the former binds to TBG, values are high in the newborn period and low in pregnancy. This observation suggests different rates of T₃S generation or metabolism in the mother and fetus. T₃S values are high in thyrotoxicosis (including patients taking suppressive doses of thyroxine), in patients receiving amiodarone therapy,⁵⁷ and in patients with nonthyroidal illness.

The mean serum concentration of T₂S in normal subjects is 0.86 ± 0.59 nmol/L, with a detection threshold between 0.17 and 0.5 nmol/L, depending on the assay. The values of T₂S were higher in hyperthyroid patients (2.2 ± 0.06) and in subjects with nonthyroidal illness (6.0 ± 1.5). It was also detectable in normal urine and amniotic fluid.⁹

Diiodotyrosine and Monoiodotyrosine

Although RIA methods have been developed for the measurement of DIT and MIT, because of limited experience, their value in clinical practice remains unknown. Early reports gave a normal mean value for DIT in the serum of normal adults of 156 ng/dL (3.6 nmol/L),⁵⁸ with a progressive decline caused by refinement of techniques to values as low as 7 ng/dL with a range of 1 to 23 ng/dL (0.02 to 0.5 nmol/L). Thus, the normal range of 90 to 390 ng/dL (2.9 to 12.7 nmol/L) for MIT is undoubtedly an overestimation. Iodotyrosine that has escaped enzymatic deiodination in the thyroid gland appears to be the principal source of DIT in serum. Iodothyronine degradation in peripheral tissues is probably a minor source of iodotyrosines because the administration of large doses of T₄ to normal subjects produces a decline rather than an increase in the serum DIT level. DIT is metabolized to MIT in peripheral tissues. Serum levels of DIT are low during pregnancy and high in cord blood. Recently, high-performance liquid chromatography (HPLC) tandem mass spectrometry was used for the measurement of iodotyrosines in urine in patients with defects in iodotyrosine deiodinase.⁵⁹

Thyronamine and 3-Iodothyronamine

Thyronamines are a novel class of endogenous signaling molecules that differ from T₄ and other T₄ derivatives owing to the absence of the carboxylate group of the β alanine side chain. Measurement of these compounds can be done by liquid chromatography and tandem mass spectrometry.⁶⁰ Only two compounds in this class, thronamine (T₀AM) and 3-tiodothyronamine (3-T₁AM), have been detected in tissues and are believed to function physiologically as thyroid hormone antagonists.^61,62 Thyronamines are isozyme-specific substrates of deiodinases.⁶⁰

Thyroglobulin

RIA methods were the methods first used routinely for measurement of Tg in serum, although other methods using immunoradiometric assay, immunochemiluminescent assay, and enzyme-linked immunosorbent assay technology have been reported and are gaining increasing popularity. They are specific and, depending on the sensitivity of the assay, capable of detecting Tg in the serum of approximately 90% of euthyroid healthy adults. When antisera are used in high dilutions, virtually no cross-reactivity with iodothyronines or iodotyrosines occurs. Results obtained from analysis of sera containing Tg autoantibodies may be inaccurate depending on the antiserum that is used.⁶³ Because of the importance of Tg measurement in the management of thyroid cancer, methods for its determination in the presence of antibodies have been devised. The simplest, and probably the most accurate, is the estimation of Tg recovery after its addition to the sample being tested.^64,65 The presence of TPO antibodies does not interfere with the Tg RIA. Despite the reliability of measurements of serum Tg, it is clear that different assay methods may result in values that are discrepant by up to 30%, even though reference preparations are available. Typically, immunochemiluminometric assay (ICMA) methods underestimate the serum Tg value, whereas RIA methods overestimate it, so it is essential that clinical decisions be based on serial measurements using the same assay.

Tg concentrations in the serum of normal adults range from less than 1 to 25 ng/mL (<1.5 to 38 pmol/L), with mean levels of 5 to 10 ng/mL.⁶⁶ On a molar basis, these concentrations of Tg are minute relative to the circulating iodothyronines: 5000-fold lower than the corresponding concentration of T₄ in serum. Values tend to be slightly higher in women than in men. In the neonatal period and during the third trimester of pregnancy, mean values are approximately fourfold and twofold higher.^67,68 They gradually decline throughout infancy, childhood, and adolescence.⁶⁹ The positive correlation between the levels of serum Tg and TSH indicates that pituitary TSH regulates the secretion of Tg.

Elevated serum Tg levels reflect increased secretory activity by stimulation of the thyroid gland or damage to thyroid tissue, whereas values below or at the level of detectability indicate a paucity of thyroid tissue or suppressed activity. Patients with acromegaly have elevated serum Tg levels, but it is unclear whether this is a direct effect of growth hormone. Tg levels in a variety of conditions affecting the thyroid gland have been reviewed 70,71 and are listed in Table 6-8.

Table 6-8

Conditions Associated With Changes in Serum Tg Concentration Listed According to the Presumed Mechanism

Increased

TSH Mediated

Acute and transient (TSH and TRH administration, neonatal period)

Chronic stimulation

Iodine deficiency, endemic goiter, goitrogens

Reduced thyroidal reserve (lingual thyroid)

TSH-producing pituitary adenoma

Resistance to thyroid hormone

TBG deficiency

Non–TSH Mediated

Thyroid stimulators

IgG (Graves’ disease)

hCG (trophoblastic disease)

Trauma to the thyroid (needle aspiration and surgery of the thyroid gland, ¹³¹I therapy)

Destructive thyroid pathology

Subacute thyroiditis

Painless thyroiditis

Postpartum thyroiditis

Abnormal release

Thyroid nodules (toxic, nontoxic, multinodular goiter)

Differentiated nonmedullary thyroid carcinoma

Abnormal clearance (renal failure)

Amiodarone-induced hyperthyroidism

Acromegaly

Cord blood

Decreased

TSH Suppression

Administration of thyroid hormone

Decreased Synthesis

Athyrosis (postoperative, congenital)

Tg synthesis defect

hCG, Human chorionic gonadotropin; IgG, immunoglobulin G; TBG, thyroxine-binding globulin; Tg, thyroglobulin; TRH, thyrotropin-releasing hormone; TSH, thyroid-stimulating hormone.

Interpretation of a serum Tg value should take into account the fact that Tg concentrations may be high under normal physiologic conditions or may be altered by drugs. Administration of iodine and antithyroid drugs increases the serum Tg level, as do states associated with hyperstimulation of the thyroid gland by TSH or other substances with thyroid-stimulating activity. This increase in serum Tg concentration is due to increased thyroidal release of Tg rather than to changes in its clearance.⁷² Administration of TRH and TSH also transiently increases the serum level of Tg.⁷³ Trauma to the thyroid gland, such as that occurring during diagnostic and therapeutic procedures, including percutaneous needle biopsy, surgery, or ¹³¹I therapy, can produce a striking although short-lived elevation in the Tg level in serum.^73,74 Pathologic processes with destructive effects on the thyroid gland also produce transient although more prolonged increases.⁷⁵ Tg is undetectable in serum after total ablation of the thyroid gland, as well as in normal people receiving suppressive doses of thyroid hormone. It is thus a useful test in the differential diagnosis of thyrotoxicosis factitia,⁷⁶ especially when transient thyrotoxicosis with low RAIU or suppression of thyroidal RAIU by iodine is an alternative possibility.

Reverse transcriptase polymerase chain reaction is a sensitive technique to measure the presence of mRNA of different genes in peripheral blood. Initial results were promising that this highly sensitive method would be useful in the management of patients with thyroid cancer ⁷⁷; however, subsequent studies have demonstrated that this measurement is of limited clinical value.^78,79

The most striking elevations in serum Tg concentrations have been observed in patients with metastatic differentiated nonmedullary thyroid carcinoma, even after total surgical and radioiodide ablation of all normal thyroid tissue. It usually persists despite full thyroid hormone suppressive therapy, suggesting excessive autonomous release of Tg by the neoplastic cells. The determination is thus of particular value in the follow-up and management of thyroid cancer metastases, particularly when they fail to concentrate radioiodide. Follow-up of such patients with sequential serum Tg determinations helps in the early detection of tumor recurrence or growth and in assessment of the efficacy of treatment. Measurement of serum Tg is also useful in patients with metastases, particularly to bone, in whom no evidence of a primary site is noted and thyroid malignancy is being considered in the differential diagnosis. On the other hand, serum Tg levels are of no value in the differential diagnosis of primary thyroid cancer because levels may be within the normal range in the presence of differentiated thyroid cancer and high in a variety of benign thyroid diseases. In fact, levels of Tg greater than 100 mg/dL can often be seen in subjects with multinodular goiters and can raise the possibility of metastatic disease. Whether early detection of recurrent thyroid cancer after initial ablative therapy could be achieved by serum Tg measurement without cessation of hormone replacement therapy is debatable because Tg secretion by the tumor is modulated by TSH and could be suppressed by the administration of thyroid hormone. Although the presence of an elevated Tg level with suppressed TSH is an indicator of probable recurrence, suppressed Tg with suppressed TSH is not a reliable indicator of the absence of recurrence. The introduction of recombinant human TSH allows stimulation of thyroid tissue for the measurement of Tg without cessation of replacement treatment.

In the early phase of subacute thyroiditis, Tg levels are high. Declining serum Tg levels during the course of antithyroid drug treatment of patients with Graves’ disease may indicate the onset of a remission.⁸⁰ Tg may be undetectable in the serum of neonates with dyshormonogenetic goiters caused by defects in Tg synthesis,⁸¹ but levels are very high in some hypothyroid infants with thyromegaly or ectopy.⁸² Measurement of serum Tg in hypothyroid neonates is useful in differentiating infants with complete thyroid agenesis from those with hypothyroidism resulting from other causes and thus in most cases obviates the need for diagnostic administration of radioiodide.

Measurement of Thyroid Hormone and Its Metabolites in Other Body Fluids and in Tissues

Clinical experience with measurement of thyroid hormone and its metabolites in body fluids other than serum and in tissues is limited for several reasons. Analyses carried out in urine and saliva do not appear to give information additional to that determined from measurements carried out in serum. Amniotic fluid, CSF, and tissues are less readily accessible for sampling. Their likely application in the future will depend on information that they could provide beyond that obtained from similar analyses in serum.

Urine

Because thyroid hormone is filtered in the urine predominantly in free form, measurement of the total amount excreted over 24 hours offers an indirect method for estimation of the free hormone concentration in serum. The 24 hour excretion of T₄ in normal adults ranges from 4 to 13 µg and from 1.8 to 3.7 µg, depending on whether total or only conjugated T₄ is measured. Corresponding normal ranges for T₃ are 2.0 to 4.0 µg and 0.4 to 1.9 µg.83–86 Striking seasonal variations have been shown for the urinary excretion of both hormones, with a nadir during the hot summer months in the absence of significant changes in serum TT₄ and TT₃. As expected, values are normal in pregnancy and in nonthyroidal illnesses and are high in thyrotoxicosis and low in hypothyroidism. The test might not be valid in the presence of gross proteinuria and impairment in renal function.

Amniotic Fluid

From week 12 of gestation onward, fetal serum concentrations of T₄ and TSH steadily rise that correlate with concentrations in the amniotic fluid ⁸⁷ and are independent of maternal concentrations.^87,88 Determination of fetal thyroid status is a clinical challenge, and although percutaneous umbilical blood sampling is technically possible, it is a demanding procedure that poses a risk for fetal bradycardia and hemorrhage. Amniocentesis, in contrast, is easier, safer, and more readily available. All iodothyronines measured in blood have also been detected in amniotic fluid. With the exception of T₃, 3,3′-T₂, and 3′-T₂, the concentration at term is lower than that in cord serum.⁸⁹ This fact cannot be fully explained by the low TBG concentration in amniotic fluid.

The TT₄ concentration in amniotic fluid averages 0.5 µg/dL (65 nmol/L) with a range of 0.15 to 1.0 µg/dL and thus is very low when compared with values in maternal and cord serum. The FT₄ concentration is, however, twice as high in amniotic fluid relative to serum. The TT₃ concentration is also low relative to maternal serum, being on average 30 ng/dL (0.46 nmol/L) in both amniotic fluid and cord serum. rT₃, on the other hand, is very high in amniotic fluid, on average 330 ng/dL (5.1 nmol/L) during the first half of gestation and declining precipitously at about the 30th week of gestation to an average of 85 ng/dL (1.3 nmol/L), which is also found at term.

A recent study sought to establish normal amniotic fluid reference intervals for TSH, total T₄, and free T₄ using automated immunoassays.⁹⁰ Results showed TSH from less than 0.1 to 0.5 mU/L, with a median of 0.1 mU/L; total T₄ from 2.3 to 3.9 µg/dL (30 to 50 nmol/L), with a median of 3.3 µg/dL (4 nmol/L); and free T₄ from less than 0.4 to 0.7 ng/dL (5 to 9 pmol/L), with a median of 0.4 ng/dL (5 pmol/L).

Cerebrospinal Fluid

T₄, T₃, and rT₃ concentrations have been measured in human CSF.91–93 The concentrations of both TT₄ and TT₃ are approximately 50-fold lower than those found in serum. However, the concentrations of these iodothyronines in free form are similar to those in serum. In contrast, the level of TrT₃ in CSF is only 2.5-fold lower than that of serum, whereas that of FrT₃ is 25-fold higher. This difference is probably due to the presence in CSF of a larger proportion of TTR, which has high affinity for rT₃. All the thyroid hormone–binding proteins that are present in serum are also found in CSF, although in lower concentrations. The concentrations of TT₄ and FT₄ are increased in thyrotoxicosis and depressed in hypothyroidism. Severe nonthyroidal illness gives rise to increased TrT₃ and FrT₃ levels.

Milk

The TT₄ concentration in human milk is on the order of 0.03 to 0.5 µg/dL.⁹⁴ Analytic artifacts were responsible for the much higher values formerly reported.⁹⁵ TT₃ concentrations range from 10 to 200 ng/dL (0.15 to 3.1 nmol/L).⁹⁶ The concentration of TrT₃ ranges from 1 to 30 ng/dL (15 to 460 pmol/L). Thus, it is unlikely that milk would provide a sufficient quantity of thyroid hormone to alleviate hypothyroidism in an infant. The serum levels of thyrotropin, T₄, free T₄, and T₃ are not significantly different between breastfed and bottle-fed babies.⁹⁷

Saliva

It has been suggested that only the free fraction of small nonpeptide hormones that circulate predominantly bound to serum proteins would be transferred to saliva and that their measurement, in this easily accessible body fluid, would provide a simple and direct means to determine their free concentration in blood. This hypothesis was confirmed for steroid hormones that are not tightly bound to serum proteins.⁹⁸ Levels of T₄ in saliva range from 4.2 to 35 ng/dL (54 to 450 pmol/L) and do not correlate with the concentration of FT₄ in serum.⁹⁹ This finding is, in part, due to the transfer of T₄ bound to small but variable amounts of serum proteins that reach the saliva.

Effusions

TT₄ measured in fluid obtained from serous cavities bears a direct relationship to the protein content and the serum concentration of T₄. Limited experience with Tg measurement in pleural effusions from patients with thyroid cancer metastatic to the lungs suggests that it might be of diagnostic value. Thyroglobulin levels are expectedly very high in fluid from thyroid nodules or cysts.

Tissues

Because the response to thyroid hormone is expressed at the cellular level via nuclear receptors, it is logical to assume that hormone concentrations in tissues should correlate best with their action. Methods for extraction, recovery, and measurement of iodothyronines from tissues have been developed, but for obvious reasons, data from thyroid hormone measurements in human tissues are limited. Preliminary work has shown that under several circumstances, hormonal levels in tissues such as liver, kidney, and muscle usually correlate with those found in serum.¹⁰⁰

Measurements of T₃ in cells most accessible for sampling in humans, namely, red blood cells, gave values of 20 to 45 ng/dL (0.31 to 0.69 nmol/L), or one-fourth those found in serum.¹⁰¹ They are higher in thyrotoxicosis and lower in hypothyroidism.

The concentrations of all iodothyronines have been measured in thyroid gland hydrolysates. In normal glands, the molar ratios relative to the concentration of T₄ are on average as follows: T₄/T₃ = 10, T₄ /rT₃ = 80, T₄/3,5′-T₂ = 1400, T₄/3,3′-T₂ = 350, T₄/3′,5′-T₂ = 1100, and T₄/3′-T₁ = 4400. Information about the content of iodothyronines in hydrolysates of abnormal thyroid tissue is limited, and the diagnostic value of such measurements has not been established.

Measurement of Tg in metastatic tissue obtained by needle biopsy may be of value in the differential diagnosis, especially when the primary site is unknown and the histologic diagnosis is not conclusive.

Tests Assessing the Effects of Thyroid Hormone on Body Tissues

Because of the ubiquitous expression of thyroid hormone receptors, thyroid hormone regulates a variety of biochemical reactions in virtually all tissues. Thus, ideally, the adequacy of hormonal supply should be assessed by tissue responses rather than by parameters of thyroid gland activity or serum hormone concentrations, which are several steps removed from the site of thyroid hormone action. Unfortunately, tissue responses (metabolic indices) are nonspecific because they are altered by a variety of physiologic and pathologic mechanisms unrelated to thyroid hormone deprivation or excess. This would indicate that thyroid function tests might not be a reliable measure of thyroid hormone action at the cellular level. These conditions include, but are not limited to, nonthyroidal illness or euthyroid sick syndrome, alterations in T₄ binding by changes in transport proteins, resistance to thyroid hormone or TSH, subclinical hypothyroidism and hyperthyroidism, central causes of hypothyroidism, and use of medications such as lithium, amiodarone, and other iodine-containing medications, and centrally acting medications. The following review of biochemical and physiologic changes mediated by thyroid hormone has a dual purpose: (1) to outline some of the changes that can be used as clinical tests in the evaluation of metabolic status, and (2) to point out the changes in various determinations commonly used in the diagnosis of a variety of nonthyroidal illnesses that might be affected by the concomitant presence of thyroid hormone deficiency or excess.

Clinical Symptom Scales

Clinical symptom scales have been developed and used in the past to aid in the diagnosis of thyrotoxicosis and hypothyroidism. A weighted score involving 19 different signs and symptoms was able to discriminate between thyrotoxic and euthyroid patients with a relatively high degree of sensitivity.¹⁰² The limitation of this scale is its basis on the presence or absence of symptoms and signs rather than on a range of degrees of their severity. A hyperthyroid scale was developed that looks at the following characteristics: nervousness, sweating, heat tolerance, hyperactivity, tremor, weakness, hyperdynamic precordium, diarrhea, appetite, and impairment of daily function.¹⁰³ Each of the symptoms or signs was graded on a scale of 0 to 4, 4 being the most severe. Newly diagnosed, untreated Graves’ disease patients had significantly higher scores compared with the same patients after treatment and euthyroid patients. Although no correlation was found between the rating scale and serum levels of T₄, total T₃, and free T₄ index, a direct relationship was found between goiter size and the rating scale, and an inverse relationship was demonstrated with age.

Clinical evaluation of hypothyroidism, in contrast, is more difficult. In 1969, Billewicz and coworkers described a diagnostic index that scores the presence or absence of various symptoms and signs of hypothyroidism for the purpose of establishing a diagnosis.¹⁰⁴ Reevaluation found that only three signs—ankle reflex, puffiness, and slow movements—had a positive predictive value greater than 90%; the rest had positive and negative predictive values of around 70% or less.¹⁰⁵ The original index was revised to exclude cold intolerance and pulse rate because these findings had a positive predictive and a negative predictive value of less than 70% in combination with an age-correcting factor, resulting in a more sensitive scale. The new scale demonstrated 62% of all overt hypothyroid and 24% of subclinical hypothyroid patients as clinically hypothyroid, compared with 42% and 6% using Billewicz’s index.

Metabolism

Basal Metabolic Rate

The basal metabolic rate (BMR) has a long history in the evaluation of thyroid function. It measures oxygen consumption under basal conditions of overnight fast and rest from mental and physical exertion. Because standard equipment for the measurement of BMR might not be readily available, the BMR can be estimated from the oxygen consumed over a timed interval by analysis of samples of expired air. The test indirectly measures metabolic energy expenditure or heat production.

Results are expressed as the percentage of deviation from normal after appropriate corrections have been made for age, gender, and body surface area. Low values are suggestive of hypothyroidism, and high values reflect thyrotoxicosis. Normal BMR ranges from negative 15% to positive 5%, most hyperthyroid patients having a BMR of positive 20% or better and hypothyroid patients commonly having a BMR of negative 20% or lower. Different clinical states are known to alter BMR. Fever, pregnancy, pheochromocytoma, adrenergic agonist drugs, cancer, congestive heart failure, acromegaly, polycythemia, and Paget’s disease of the bone are known to increase the BMR. Obesity, starvation or anorexia, hypogonadism, adrenal insufficiency, Cushing’s syndrome, immobilization, and sedative drugs are known to decrease the BMR.

Metabolic Markers

Plasma homocysteine concentrations have been shown to increase in hypothyroidism and decrease in hyperthyroidism, indicating that free T₄ is an independent determinant of total homocysteine concentrations.106–108 A longitudinal study of hyperthyroid and hypothyroid patients over 12 months of treatment showed that serum homocysteine levels started at higher levels in hypothyroid patients and at lower levels in hyperthyroid patients; with treatment, the values of both patient groups approached the same values. Lower folate levels and a lower creatinine clearance in hypothyroidism and a higher creatinine clearance in hyperthyroidism only partially explain these changes in homocysteine. The association of elevated homocysteine levels in subclinical hypothyroidism is less clear.¹⁰⁹ Free fatty acids in serum are higher in hyperthyroid patients than in controls or in hypothyroid patients and might be a marker of lipolysis.^110,111 Serum glycerol levels are significantly increased in hyperthyroid patients and are lower in hypothyroid patients versus controls. Ketone bodies are increased in hyperthyroidism.¹¹²

Deep Tendon Reflex Relaxation Time (Photomotogram)

A delay in the relaxation time of the deep tendon reflexes, visible to the experienced eye, occurs in hypothyroidism. Several instruments have been devised to quantitate various phases of the Achilles tendon reflex. Although normal values vary according to the phase of the tendon reflex measured, the apparatus used, and individual laboratory standards, the approximate adult normal range for the half-relaxation time is 230 to 390 msec. Diurnal variation, differences with gender, and changes with age, cold exposure, fever, exercise, obesity, and pregnancy have been reported. However, the main reason for failure of this test as a diagnostic measure of thyroid dysfunction is the large overlap with values obtained in euthyroid individuals and alterations caused by nonthyroidal illnesses.¹¹³

Tests Related to Cardiovascular Function

Thyroid hormone affects the heart through regulation of cardiac gene expression as well as through nongenomic means. Evidence of thyroid hormone–regulated cardiac gene expression comes from in vivo studies of transcription of the cardiac myocyte gene α-myosin heavy chain (α-MHC). Myocyte genes that are positively regulated by thyroid hormones include α-MHC, sarcoplasmic reticulum Ca²⁺-adenosine phosphatase (ATPase SERCA), and the voltage-gated potassium channels Kv1.4, Kv4.2, and Kv4.3.114,115 These are all critical determinants of contractile activity and are downregulated in hypothyroidism. Genes that are negatively regulated by thyroid hormone include the β-myosin heavy chain (β-MHC) and phospholamban, which regulates contractile function through modulation of calcium cycling. These genes are upregulated in hypothyroidism.

Thyroid hormone can also affect the myocardium through nongenomic actions. Changes primarily involve membrane ion channels and ion pumps, but other possible extranuclear changes involving cell-surface proteins, signal transduction, and intracellular protein trafficking, myocardial contractility and metabolism, vascular smooth muscle, and myocardial mitochondria have been described.¹¹⁶ Nongenomic actions of thyroid hormone, primarily T₃, on the plasma membrane of myocytes include (1) stimulation of the Na⁺/H⁺ antiporter through effects on the activity of protein kinase C,¹¹⁷ (2) stimulation of Ca²⁺-adenosine triphosphatase (ATPase) activity through activation of phospholipase C (PLC)¹¹⁸ and presence of calmodulin, (3) prolonged activation of the Na⁺ current through effects on protein kinase C activity, and (4) an increase in the inward rectifying K⁺ current, which may be mediated via G protein–coupled receptors.¹¹⁹ These extranuclear nontranscriptional effects of thyroid hormones on the performance characteristics of the ion channels in the heart result in changes in intracellular levels of calcium and potassium, which can increase inotropy and chronotropy. Noninvasive measures of cardiovascular hemodynamics using electrocardiogram, echocardiogram, and Doppler parameters have been found to be a sensitive measure of T₃ action on the cardiac and vascular smooth muscle cells.¹²⁰ Standardized measures of cardiac systolic function include (1) the pre-ejection period (PEP), which is the time from the QRS complex onset to the opening of the aortic valve, and (2) left ventricular ejection time, or the time from the opening of the aortic valve to the end of ventricular systole. Two other highly reproducible measures that are obtained via two-dimensional echocardiogram are the isovolumetric contraction time (ICT) and the isovolumetric relaxation time (IVRT), which are measures of early systole and of diastolic function, respectively. In hyperthyroidism secondary to Graves’ disease and multinodular goiter, increased systolic function with shortened PEP, left ventricular ejection time (LVET), and ICT; reduced systemic vascular resistance; and supranormal diastolic dysfunction with shortened left ventricular relative thickness (LVRT) and rate of blood flow across the mitral valve have been demonstrated. In hypothyroidism, all measures of cardiac contractile performance are impaired. Cardiac function improved as soon as 24 hours after treatment and returned to normal levels by 1 week in a study using intravenous T₃ for 1 week in hypothyroid patients.¹²¹ The geometry, cardiac function, and oxidative metabolism have also been assessed in hypothyroid patients by positron emission tomography (PET) and magnetic resonance imaging (MRI). Ejection fraction and myocardial efficiency were derived from the imaging measurements and were found to be decreased in hypothyroidism and improved with thyroid hormone treatment.¹²² In hyperthyroidism, heart rate and cardiac output are expectedly higher. Peripheral vascular resistance is reduced. Differences in blood pressure, stroke volume, and ventricular mass are not observed.

Neurobehavioral Markers of Thyroid Hormone Action

A variety of neuropsychiatric scales have also been used to evaluate hyperthyroid and hypothyroid patients and their response to treatment,¹²³ including a specific thyroid symptom scale for hypothyroid patients.¹²⁴ A study of hyperthyroid patients found them to have abnormal scores on neuropsychological tests (IQ, memory, and attention span) that improved somewhat with propranolol therapy and more with antithyroid drug therapy. Disturbances in cognition in hypothyroidism include inattentiveness, inability to concentrate, slowing of thought processes and speech, inability to calculate and to understand complex questions, and alterations in perception.¹²⁵ In addition, memory deficits, particularly for recent events, have been found to occur in hypothyroidism. Furthermore, hypothyroidism in nondemented older adults is associated with impairments in learning, word fluency, visual-spatial abilities, attention, visual scanning, and psychomotor function. Several studies evaluating the effects of T₄ and T₃ treatment on neurocognitive function and psychiatric symptoms with varying outcomes, however, demonstrate that the underlying pathophysiologic mechanisms of thyroid hormone in the brain still need further elucidation.^126,127 The metabolic consequences of hypothyroidism have been studied with the use of imaging techniques such as ³¹P nuclear magnetic spectroscopy, which demonstrated an increase in the phosphocreatinine/inorganic phosphate ratio after treatment for acute hypothyroidism.¹²⁸ A more recent study used PET to correlate the regional cerebral blood flow and cerebral glucose metabolism with the mental state in patients. Investigators demonstrated a generalized decrease in regional cerebral blood flow by 23.4% and in cerebral glucose metabolism by 12.1% and no specific local defects.¹²⁹

Miscellaneous Biochemical and Physiologic Changes Related to The Action of Thyroid Hormone on Peripheral Tissues

Thyroid hormone affects the function of a variety of peripheral tissues. Thus, hormone deficiency or excess can alter a number of the determinations that are used in the diagnosis of illnesses unrelated to thyroid hormone dysfunction. Knowledge of the determinations that may be affected by thyroid hormone is important in the interpretation of laboratory data.

Measurement of Substances Absent in Normal Serum

Tests that measure substances present in the circulation only under pathologic circumstances do not provide information on the level of thyroid gland function. They are of value in establishing the cause of the hormonal dysfunction or thyroid gland pathology.

Thyroid Autoantibodies

In clinical practice, the antibodies that are most commonly measured are directed against Tg or thyroid cell microsomal proteins. The latter is principally represented by thyroperoxidase (TPO). Immunoassays have been developed with the use of purified and recombinant TPO.130–132 Other circulating immunoglobulins, which are used less frequently as diagnostic markers, are those directed against a colloid antigen, T₄, and T₃. Immunoglobulins that have the property of stimulating the thyroid gland will be discussed in the next section.

A variety of techniques have been developed for the measurement of Tg and microsomal antibodies. These procedures include a competitive binding radioassay, complement fixation reaction, tanned red cell agglutination assay, Coon’s immunofluorescent technique, and enzyme-linked immunosorbent assay. Although the competitive binding radioassay is a sensitive test, agglutination methods combine sensitivity and simplicity and have now largely superseded other methods. Current commercial kits use synthetic gelatin beads rather than red cells.

In the assay of Tg and TPO antibodies by hemagglutination, particulate material is coated with human Tg or solubilized thyroid microsomal proteins and is exposed to serial dilutions of the patient’s serum. Agglutination of the coated particulate material occurs in the presence of antibodies that are specific to the antigen attached to their surface. To detect false-positive reactions, it is important to include a blank for each sample consisting of uncoated particles. Because of the common occurrence of a prozone or blocking phenomenon, it is necessary to screen all serum samples through at least six consecutive twofold dilutions. Results are expressed in terms of the highest serum dilution, or titer, that shows persistent agglutination. The presence of immune complexes, particularly in patients with high serum Tg levels, may mask the presence of Tg antibodies. Assays have been developed for the measurement of such Tg–anti-Tg immune complexes.¹³³

Normally, the test response is negative, but results may be positive in up to 10% of the adult population. The frequency of positive test results is higher in women and with advancing age. The presence of thyroid autoantibodies in the apparently healthy population is thought to represent subclinical autoimmune thyroid disease rather than false-positive reactions. Nonetheless, it is difficult to compare results from such studies because some laboratories that use agglutination methods report low titers (1/10 to 1/40) as positive. It is important in reporting values that a method-specific normal range be used, and that assays be calibrated against internationally available reference preparations. The availability of such preparations allows the reporting of results in international units. The importance of using internationally available reference ranges is confirmed by preliminary evidence that suggests that TPO and Tg antibody testing might be less reliable in certain ethnic groups. TPO antibodies are detectable in approximately 95% of patients with Hashimoto’s thyroiditis and 85% of those with Graves’ disease, irrespective of the functional state of the thyroid gland. Similarly, Tg antibodies are positive in about 60% and 30% of adult patients with Hashimoto’s thyroiditis and Graves’ disease, respectively. Tg antibodies are less frequently detected in children with autoimmune thyroid disease. Although higher titers are more common in Hashimoto’s thyroiditis, quantitation of the antibody titer carries little diagnostic implication. The tests are of particular value in the evaluation of patients with atypical or selected manifestations of autoimmune thyroid disease (ophthalmopathy and dermopathy). Positive antibody titers are predictive of postpartum thyroiditis.¹³⁴ Low antibody titers occur transiently in some patients after an episode of subacute thyroiditis,¹³⁵ presumably caused by antigen exposure. No increased incidence of thyroid autoantibodies is seen in patients with multinodular goiter, thyroid adenomas, or secondary hypothyroidism. In some patients with Hashimoto’s thyroiditis and undetectable thyroid autoantibodies in their serum, intrathyroidal lymphocytes have been demonstrated to produce TPO antibodies.

The development of sensitive radioassays and quantitative enzyme-linked immunoassays has allowed measurement of the concentration of antibodies to TPO and Tg in absolute terms. Such methods provide sensitive, precise, and antigen-specific means of revealing quantitative fluctuations in autoantibody concentrations.

Other antibodies that are directed against thyroid components (such as NIS) or other tissues have been detected in the serum of some patients with autoimmune thyroid disease. They do not exert a blocking effect on NIS,¹³⁶ and their diagnostic value has not been fully evaluated. Circulating antibodies that are capable of binding T₄ and T₃ have also been demonstrated in patients with autoimmune thyroid disease and might interfere with the measurement of T₄ and T₃ by immunometric techniques. Antibodies reacting with nuclear components, which are not tissue specific, and with cellular components of parietal cells and adrenal, ovarian, and testicular tissue are more commonly encountered in patients with autoimmune thyroid disease.¹³⁷ Their presence reflects the frequency of coexistence of several autoimmune disease processes in the same patient.

Thyroid-Stimulating Immunoglobulins

A large number of names have been given to tests that measure abnormal gamma globulins present in the serum of some patients with autoimmune thyroid disease, in particular Graves’ disease.¹³⁸ The interaction of these unfractionated immunoglobulins with thyroid follicular cells usually results in global stimulation of thyroid gland activity and only rarely causes inhibition. It has been recommended that these assays all be called TSH receptor antibodies with the phrase “measured by … assay” to identify the type of method that was used for their determination. The tests will be described under three general categories: (1) those measuring thyroid-stimulating activity by using in vivo or in vitro bioassays, (2) tests based on competition of the abnormal immunoglobulin with binding of TSH to its receptor, and (3) measurement of the thyroid growth–promoting activity of immunoglobulins. Tests use both human and animal tissue material or cell lines.

Thyroid Stimulation Assays

The earliest assays used various modifications of the McKenzie mouse bioassay.139,140 The abnormal gamma globulin with TSH-like biological properties has relatively longer in vivo activity, hence its name, long-acting thyroid stimulator (LATS). The assay measures the LATS-induced release of thyroid hormone from the mouse thyroid gland prelabeled with radioiodide. The presence of LATS in serum is pathognomonic of Graves’ disease. However, depending on assay sensitivity, a variable percentage of untreated patients will show a positive LATS response. LATS activity may be found in the serum of patients with Graves’ disease even in the absence of thyrotoxicosis. Although it is more commonly present in patients with ophthalmopathy, especially when accompanied by pretibial myxedema,¹⁴¹ LATS activity does not appear to correlate with the presence of Graves’ disease, its severity, or the course of complications. LATS crosses the placenta and may be found transiently in newborns of mothers possessing the abnormal gamma globulin.¹⁴²

Attempts to improve the ability to detect thyroid-stimulating antibodies (TSAbs) in autoimmune thyroid disease led to the development of several in vitro assays using animal as well as human thyroid tissue. The ability of the patient’s serum to stimulate endocytosis in fresh human thyroid tissue is measured by direct count of the intracellular colloid droplets formed. When such a technique is used, human thyroid stimulator activity has been demonstrated in serum samples from patients with Graves’ disease that were devoid of LATS activity measured by the standard mouse bioassay.¹⁴³ TSAbs can be detected by measuring the accumulation of cyclic adenosine monophosphate (cAMP) or the stimulation of adenylate cyclase activity in human thyroid cell cultures and thyroid plasma membranes, respectively. Accumulation of cAMP in the cultured rat thyroid cell line FRTL5 has also been used as an assay for TSAb.¹⁴⁴ Stimulation of release of T₃ from human and porcine thyroid slices is another form of in vitro assay for TSAb. An in vitro bioassay using a cytochemical technique depends on the ability of thyroid-stimulating material to increase lysosomal membrane permeability to a chromogenic substrate, leucyl-β-naphthylamide, which then reacts with the enzyme naphthylamidase. Quantitation is by scanning and integrated microdensitometry.¹⁴⁵

Cloning of the TSH receptor led to the development of an in vitro assay for TSAb in cell lines that express the recombinant TSH receptor.146–148 This assay, based on the generation of cAMP, is specific for the measurement of human TSH receptor antibodies that have thyroid-stimulating activity and thus contrasts with assays based on binding to the TSH receptor (see below), which cannot distinguish between antibodies with thyroid-stimulating and TSH-blocking activity. Accordingly, the recombinant human TSH receptor assay measures antibodies that are relevant to the pathogenesis of autoimmune thyrotoxicosis and is more sensitive than the formerly used TSAb assays.¹⁴⁹ For example, 94% of serum samples were positive for TSAb compared with 74% when the same samples were assayed by using FRTL5 cells.¹⁵⁰

One of the major drawbacks with the bioassays mentioned above is that they are unsuitable for use as routine laboratory tools. This problem has been largely circumvented, however, with the advent of a luminescence-linked bioassay for TSAbs. This assay uses Chinese hamster ovary cells stably transfected with the human TSH receptor and a cAMP-dependent luciferase reporter. With the use of an international TSAb standard, the assay responded in a dose-dependent manner, with 10 mIU/mL of standard generating a relative light unit score of greater than 10. The authors also reported that the assay was modified to allow use in a 96-well plate format, thereby permitting automated measurement of relative luminosity in a large number of samples.

Thyrotropin-Binding Inhibition Assays

The principle of binding inhibition assays dates to the discovery of another class of abnormal immunoglobulins in patients with Graves’ disease: those that neutralize the bioactivity of LATS tested in the mouse.¹⁵¹ This material, known as the LATS protector, is species specific; it has no biological effect on the mouse thyroid gland but is capable of stimulating the human thyroid.¹⁵² The original assay was cumbersome, which limited its clinical application.

Techniques that are used currently, which may be collectively termed radioreceptor assays, are based on competition of the abnormal immunoglobulins and TSH for a common receptor-binding site on thyroid cells. The test is akin in principle to the radioligand assays, in which a natural membrane receptor takes the place of the binding proteins or antibodies. Various sources of TSH receptor are used, including human thyroid cells, their particulate or solubilized membrane, and cell membranes from porcine thyroids or guinea pig fat cells or recombinant human TSH receptor expressed in mammalian cells. Because the assays do not directly measure thyroid-stimulating activity, the abnormal immunoglobulins determined have been given a variety of names, such as thyroid-binding inhibitory immunoglobulins or antibodies and thyrotropin-displacing immunoglobulins. This type of assay has indicated that not all the antibodies that are detected stimulate the thyroid, and some are inhibitory. Even with modern techniques, the presence of inhibitory antibody is less sensitive and specific for Graves’ disease than is the presence of stimulatory antibody activity.¹⁵³ The stimulatory and inhibitory effects can be differentiated only by functional assays, which typically measure the production of cAMP.

Thyroid Growth–Promoting Assays

Assays have also been developed that measure the growth-promoting activity of abnormal immunoglobulins. One such assay is based on the staining of nuclei from guinea pig thyroid cells in S phase by the Feulgen reaction.¹⁵⁴ Another assay measures the incorporation of ³H-thymidine into DNA in FRTL cells.¹⁵⁵ Whether the thyroid growth–stimulating immunoglobulins that are measured by these assays represent a population of immunoglobulins distinct from those with stimulatory functional activity remains a subject of active debate.

Clinical Applications: Measurement of abnormal immunoglobulins that interact with thyroid tissue by any of the methods described above is not indicated as a routine diagnostic test for Graves’ disease. It is useful, however, in a few selected clinical conditions: (1) in the differential diagnosis of exophthalmos, particularly unilateral exophthalmos, when the origin of this condition is otherwise not apparent; the presence of thyroid-stimulating immunoglobulins would obviate the necessity to undertake more complex diagnostic procedures described elsewhere ¹⁵⁶; (2) in the differential diagnosis of pretibial myxedema and other forms of dermopathy when the cause is unclear and it is imperative that the cause of the skin lesion be ascertained; (3) in the differentiation of Graves’ disease from toxic nodular goiter when both are being considered as the possible cause of thyrotoxicosis, when other tests such as thyroid scanning and thyroid autoantibody tests have been inconclusive, and particularly when such a distinction would play a role in determining the course of therapy; (4) when nonautoimmune thyrotoxicosis is suspected in a patient with hyperthyroidism and diffuse or nodular goiter; (5) in Graves’ disease during pregnancy, when high maternal levels of TSAb are a warning for the possible occurrence of neonatal thyrotoxicosis; and (6) in neonatal thyrotoxicosis, where serial TSAb determinations showing a gradual decrease may be helpful in distinguishing between intrinsic Graves’ disease in an infant and transient thyrotoxicosis resulting from passive transfer of maternal TSAb. Some investigators have found the persistence of TSAbs to be predictive of relapse of Graves’ thyrotoxicosis after a course of antithyroid drug therapy.¹⁵⁷

Other Substances with Thyroid-Stimulating Activity

Hyperthyroidism develops in some patients with trophoblastic disease as a result of the production and release of a thyroid stimulator that has been termed molar or trophoblastic thyrotropin or big placental TSH.¹⁵⁸ It is likely that the thyroid-stimulating activity in patients with trophoblastic disease is entirely due to the presence of high levels of human chorionic gonadotropin (hCG).¹⁵⁹ Thus, an RIA for hCG can be useful in the differential diagnosis of thyroid dysfunction, although the clinical status of the patient is likely to suggest this cause.

The thyroid-stimulating activity of hCG can be enhanced by mutations in the TSH receptor that increase its affinity for this placental hormone. The consequence is thyrotoxicosis limited to pregnancy and associated with hyperemesis gravidarum in women.¹⁶⁰

Exophthalmos-Producing Substance

A variety of tests have been developed for measuring exophthalmogenic activity in serum. Although great uncertainty still exists regarding the pathogenesis of thyroid-associated eye disease, the role of the immune system appears to be central. Exophthalmogenic activity has also been detected in IgG fractions of some patients with Graves’ ophthalmopathy. Autoantibodies directed toward a 64 kilodalton eye muscle protein have been identified in 73% of patients with active thyroid-associated ophthalmopathy,161–163 although the role of these antibodies in the diagnosis and management of the ophthalmopathy is at present unclear.

Tests of Cell-Mediated Immunity

Delayed hypersensitivity reactions to thyroid antigens are present in autoimmune thyroid diseases. Cell-mediated immunity can be measured in several ways: (1) the migration inhibition test, which measures the inhibition of migration of sensitized leukocytes when exposed to the sensitizing antigen; (2) the lymphotoxic assay, which measures the ability of sensitized lymphocytes to kill target cells when exposed to the antigen; (3) the blastogenesis assay, which scores the formation of blast cells after exposure of lymphocytes to a thyroid antigen; and (4) thymus-dependent (T) lymphocyte subset quantitation using monoclonal antibodies.

Anatomic and Tissue Diagnoses

The purpose of the procedures described in this section is to evaluate the anatomic features of the thyroid gland, to localize and determine the nature of abnormal areas, and eventually to provide a pathologic or tissue diagnosis.

Thyroid Scintiscanning

Normal and abnormal thyroid tissue can be externally imaged by three scintiscanning methods: (1) with radionuclides that are concentrated by normal thyroid tissues, such as iodide isotopes and ^99mTc given as the pertechnetate ion; (2) by administration of radiopharmaceutical agents that are preferentially concentrated by abnormal thyroid tissues; and (3) by fluorescent scanning, which uses an external source of ²⁴¹Am and does not require the administration of radioactive material. Each has specific indications, advantages, and disadvantages.

The physical properties, dosages, and radiation delivered by the most commonly used radioisotopes are listed in Table 6-1. The choice of scanning agent depends on the purpose of the scan, the age of the patient, and the equipment available. Radioiodide scans cannot be performed in patients who have recently ingested iodine-containing compounds. ¹²³I and ^99mTcO₄⁻ are the radionuclides of choice because of the low radiation exposure. The radioisotope ¹³¹I is still used for the detection of functioning metastatic thyroid carcinoma by total body scanning.

Radioiodide and ^99mTc Scans

^99mTcO₄⁻ is concentrated, and all iodide isotopes are concentrated and bound by thyroid tissue. Depending on the isotope that is used, scans are carried out at different times after administration: 20 minutes for ^99mTcO₄⁻; 4 or 24 hours for ¹²³I; 24 hours for ¹²⁵I and ¹³¹I; and 48, 72, and 96 hours when ¹³¹I is used in the search for metastatic thyroid carcinoma. The appearance of the normal thyroid gland on scan may be best described as a narrow-winged butterfly. Each “wing” represents a thyroid lobe, which in the adult measures 5 ± 1 cm in length and 2.3 ± 0.5 cm in width. Common variants include the absence of a connecting isthmus, a large isthmus, asymmetry between the two lobes, and trailing activity extending to the cricoid cartilage (pyramidal lobe). The latter is more commonly found in conditions associated with diffuse thyroid hyperplasia. Occasionally, collection of saliva in the esophagus during ^99mTcO₄⁻ scanning may simulate a pyramidal lobe, but this artifact can be eliminated by drinking water.

Indications for scanning are listed in Table 6-9. In clinical practice, scans most often are requested for evaluation of the functional activity of solitary nodules. However, owing to the widespread use of fine-needle aspiration in the initial evaluation of thyroid nodules, this indication has been less used, and scans generally are obtained to search for abnormal uptake of iodine outside the thyroid bed in patients with thyroid cancer and congenital hypothyroidism. Normally, the isotope is homogeneously distributed throughout both lobes of the thyroid gland. This diffuse distribution occurs in the enlarged gland of Graves’ disease and may be seen in Hashimoto’s thyroiditis. A mottled appearance may be noted in Hashimoto’s thyroiditis and can occasionally be seen in Graves’ disease, especially after therapy with radioactive iodide. Irregular areas of relatively diminished and occasionally increased uptake are characteristic of large multinodular goiters. The traditional nuclear medicine jargon classifies nodules as “hot,” “warm,” and “cold,” according to their isotope-concentrating ability relative to the surrounding normal parenchyma. Hot, or hyperfunctioning, nodules are typically benign, although the presence of malignancy has been reported.^164,165 In a large series of patients in whom thyroid scans were undertaken and subsequent surgery was performed irrespective of the scan result, 84% were cold nodules, 11% were warm nodules, and 6% were hot nodules. A histopathologic diagnosis of thyroid cancer was made in 16% of cold, 9% of warm, and 4% of hot nodules.^166,167 These data demonstrate that cold nodules are most likely to harbor malignancy but that most cold nodules are benign and, furthermore, the finding of a hot nodule on thyroid isotope scanning does not exclude the presence of malignancy. Occasionally, a nodule that is functional on a ^99mTcO₄⁻ scan will be found to be cold on an iodine scan; this pattern is found with both benign and malignant nodules.

Table 6-9

Indications for Radionuclide Scanning

Detection of anatomic variants and search for ectopic thyroid tissue (thyroid hemiagenesis, lingual thyroid, struma ovarii)

Diagnosis of congenital athyrosis

Determination of the nature of abnormal neck or chest (mediastinal) masses

Evaluation of solitary thyroid nodules (functioning or nonfunctioning)

Evaluation of thyroid remnants after surgery

Detection of functioning thyroid metastases

Evaluation of focal functional thyroid abnormalities (suppressed or nonsuppressible tissue)

Thyroid isotope scans are of particular value in identifying autonomous thyroid nodules because the remainder of the activity in the gland is suppressed. This can have important therapeutic implications in that hyperthyroid subjects with a hot thyroid nodule would more appropriately be treated with ¹³¹I than with long-term use of antithyroid medications.

A search for functioning thyroid metastases is best accomplished by using 2 to 4 mCi of ¹³¹I after ablation of normal thyroid tissue and cessation or reduction of the amount of hormone to allow TSH to increase above the upper limit of normal. In general, two methods are used for total body scanning of thyroid cancer patients following surgery and initial radioactive iodine ablation. In one instance, the patient is made hypothyroid by discontinuing the regular dose of levothyroxine (L-T₄) and taking 25 µg liothyronine (L-T₃) twice daily for 2 weeks followed by no exogenous thyroid hormone treatment. When the TSH is 20 mU/L or greater, then a 72-hour total body scan is done with 2 mCi ¹³¹I. Alternatively, the patient can be made moderately hypothyroid by taking the routine dose of L-T₄ every other day.¹⁶⁸ This will result in an elevated TSH of 20 mU/L or greater without the severe symptoms of hypothyroidism experienced in the previous method. Recombinant human TSH (r-hTSH, Thyrogen) can be used as an alternative to thyroid hormone withdrawal with approximately the same overall sensitivity and specificity for detection of thyroid metastasis. According to the Thyrogen protocol for scanning, 0.9 mg of r-hTSH is given via intramuscular injections on day 1 and day 2. On day 3, 4 mCi of ¹³¹I is given, and a total body scan is performed on day 5, 48 hours after isotope ingestion. Blood is obtained for TSH and Tg measurements on day 1 and day 5.^169,170

Uptake is also found outside the thyroid gland in patients with lingual thyroids and in the rare ovarian dermoid tumor containing functioning thyroid tissue.

The scan can be used as an adjunct during TSH stimulation and T₃ suppression tests to localize suppressed normal thyroid tissue and autonomously functioning areas, respectively (see below). Applications other than those listed in Table 6-9 are of doubtful benefit and are rarely justified in view of the radiation exposure, expense, and inconvenience. ¹²³I single-photon emission computed tomography (CT) may also be useful in the evaluation of thyroid abnormalities.

¹⁸Fluorodeoxyglucose Positron Emission Tomography Scanning

Some patients with thyroid cancer metastases do not concentrate ¹³¹I, even when it is given in therapeutic doses.¹⁷¹ Although other imaging modalities such as ultrasound, CT scanning, and nuclear magnetic resonance imaging can be used to detect the locations of the metastases, none are 100% specific or sensitive (see below). Several studies have shown that in differentiated thyroid carcinoma, fluorodeoxyglucose (FDG) PET can be used to detect recurrence or metastases with a high degree of sensitivity (80% to 90%).^172–174 In addition, the combination of FDG PET and CT scanning can increase the yield of either modality alone.¹⁷⁴

Other Isotope Scans

Because most test procedures, short of direct microscopic examination of thyroid tissue, fail to detect thyroid malignancy with any degree of certainty, efforts have been made to find other radioactive materials that would possibly be of diagnostic use. Several such agents that are concentrated by metabolically active tissues, irrespective of whether they have iodide-concentrating ability, have been tried. However, despite claims to the contrary, either they have had only limited value or their diagnostic usefulness has not been fully evaluated. These agents include ⁷⁵Se-methionine, ¹²⁵Ce, ⁶⁷Ga-citrate, ³²P-pyrophosphate, ^99mTc, and ²⁰¹Th.¹⁷⁵

Scanning with ¹³¹I-labeled anti-Tg for the detection of occult metastatic thyroid malignancy that fails to concentrate ¹³¹I showed early promising results.¹⁷⁶ However, the procedure has not proved clinically useful.

Ultrasonography

Ultrasonography is used to outline the thyroid gland and to characterize lesions differing in density from the surrounding tissue. The technique differentiates interphases of different acoustic densities by using sound frequencies in the megahertz range that are above the audible range. A transducer fitted with a piezoelectric crystal produces and transmits the signal and receives echo reflections. Interfaces of different acoustic densities reflect dense echoes, liquid transmits sound without reflections, and air-filled spaces do not transmit the ultrasound.¹⁷⁷

One of the most useful applications of the ultrasonogram is in the differentiation of solid from cystic lesions. Purely cystic lesions are entirely sonolucent, whereas solid lesions produce multiple echoes as a result of multiple sonic interphases. Many lesions, however, are mixed (solid and cystic) and are termed complex lesions. Some tumors can have the same acoustic characteristics as the surrounding normal tissue, thus escaping echographic detection. True cystic lesions are usually benign, although most large thyroid nodules contain cystic areas. Carcinomas may also undergo infarction, which causes degenerative cystic change within their substance. In one large series of thyroid nodules, 69% were found to be solid, and 21% of these were subsequently shown to be malignant. Similarly, of the 19% of cystic lesions, 7% were malignant, and malignancy was also detected in 12% of the mixed lesions.166,167 Thus, solid nodules, as determined by ultrasound scanning, were most likely to be malignant, but most were benign, and a similar proportion of cystic nodules were also malignant. Ultrasound scanning of thyroid nodules is poorly sensitive and specific as an indicator of the presence of malignancy in a thyroid mass.

Although high-resolution ultrasonography can detect thyroid nodules on the order of a few millimeters, lesions must be larger than 0.5 cm to allow differentiation between solid and cystic structures. A sonolucent pattern is frequently noted in glands with Hashimoto’s thyroiditis, but this pattern has also been described in multinodular glands and in patients with Graves’ disease.

Because sonography localizes the position and the depth of lesions, the procedure has been used to successfully guide the needle during aspiration biopsy. In complex lesions, sonographic guiding ensures sampling from the solid portion of the nodule. With experience and proper calibration, sonography can be used for estimation of thyroid gland size. Several recent reports have described treatment for toxic nodules provided by the injection of alcohol under sonographic guidance. Although ultrasonography has found virtually the same applications as scintiscanning, claims that the former may differentiate benign from malignant lesions are unfounded. Also, ultrasonography cannot be used for the assessment of substernal goiters because of interference from overlying bone.

The procedure is simple and painless and, at the frequencies of sound that are used, does not produce tissue damage. Because it does not require the administration of isotopes, it can be safely used in children and during pregnancy. Also, because the procedure is independent of iodine-concentrating mechanisms, it is valuable in the study of suppressed glands.

X-ray Procedures

A simple x-ray film of the neck and upper part of the mediastinum can provide valuable information regarding the location, size, and effect of a goiter on surrounding structures. Radiographs may show an asymmetrical goiter, an intrathoracic extension of the gland, and displacement or narrowing of the trachea. If any suggestion of posterior extension of the mass is seen, it is useful to take films during the swallow of x-ray contrast material. The soft tissue x-ray technique may disclose calcium deposits. Large deposits in flakes or rings are typical of an old multinodular goiter, whereas foci of finely stippled flecks of calcium are suggestive of papillary adenocarcinoma.

Information not related to anatomic abnormalities of the thyroid gland may be obtained from x-ray studies. In children with a history of hypothyroidism, an x-ray film of the hand to determine bone age could aid in estimating the onset and duration of thyroid dysfunction. Hypothyroidism leads to retardation in bone age and in infants produces dense calcification of epiphyseal plates most easily seen at the distal end of the radius. Long-standing myxedema produces pituitary hypertrophy, which especially in children, but also in adults, causes enlargement of the sella turcica on imaging of the pituitary region.

CT and MRI provide useful information on the location and architecture of the thyroid gland, as well as its relationship to surrounding tissues. An important application of CT is the assessment and delineation of obscure mediastinal masses and large substernal goiters.¹⁷⁸ The necessity to infuse iodine-containing contrast agents limits the application of CT in patients who are being considered for radioiodide therapy. CT and MRI have found firm application in another area of thyroid diseases: the evaluation of ophthalmopathy and mediastinal masses.

Other Procedures

A barium swallow may be useful in evaluating impingement of a goiter on the esophagus, whereas a flow-volume loop ¹⁷⁹ may be useful in allowing quantitative documentation of functional impingement on the upper part of the airway.

Biopsy of The Thyroid Gland

Histologic examination of thyroid tissue for diagnostic purposes requires some form of an invasive procedure. The biopsy procedure depends on the intended type of microscopic examination. Core biopsy for histologic examination of tissue with preservation of architecture is obtained by closed needle or open surgical procedures; aspiration biopsy is performed to obtain material for cytologic examination.

Core Biopsy

Closed core biopsy is an office procedure that is carried out with the patient under local anesthesia. A large (about 15 gauge) cutting needle of the Vim-Silverman type is most commonly used. The needle is introduced under local anesthesia through a small skin nick, and firm pressure is applied over the puncture site for 5 to 10 minutes after withdrawal of the needle. In experienced hands, complications are rare, but they may be serious and can include transient damage to the laryngeal nerve, puncture of the trachea, laryngospasm, jugular vein phlebitis, and bleeding. With the improvement in cytology based on fine-needle aspiration (see below), the use of core and open biopsy has been virtually abandoned.

Percutaneous Fine-Needle Aspiration

The development of more sophisticated staining techniques for cytologic examination, the realization that fear of tumor dissemination along the needle track was not well founded, and especially the high diagnostic accuracy of the technique are responsible for the increasing popularity of percutaneous fine-needle aspiration (FNA).

The procedure is exceedingly simple and safe. The patient lies supine, with the neck hyperextended by placing a small pillow under the shoulders. Local anesthesia is not usually required. The skin is prepared with an antiseptic solution. The lesion, fixed between two gloved fingers, is penetrated with a fine (22 to 27 gauge) needle attached to a syringe. Suction is then applied while the needle is moved within the nodule. A nonsuction technique using capillary action has also been developed. The small amount of aspirated material, usually contained within the needle or its hub, is applied to glass slides and is spread. Some slides are air-dried and others are fixed before staining. Because biopsy of small nodules may be technically more difficult, the use of ultrasound to guide the needle is recommended. It is important that the slides be properly prepared, stained, and read by a cytologist who is experienced in the interpretation of material from thyroid gland aspirates. Newer molecular approaches have been used to analyze FNA material to help distinguish benign from malignant pathologies, although these techniques are currently in development stages.

The yield of false-positive and false-negative results is variable from one center to another, but both are acceptably low. Various centers have reported that the accuracy of this technique in distinguishing benign from malignant lesions may be as high as 95%. The false-positive rate is about 2%, and the false-negative rate is about 1% to 6%, cystic papillary carcinoma being the most common cause of false negative. Patients with an initial FNA that demonstrates frankly malignant or suspicious cytologic findings (including follicular neoplasms; see below) on first examination are immediately referred for surgery. Some clinicians recommend that repeat FNA be performed 6 to 12 months after the initial biopsy to reduce the risk of false-negative results. In one study, the second FNA changed the diagnosis in 7% of patients, and an additional four carcinomas were detected.

In one clinic in which the procedure is used routinely, the number of patients operated on decreased by one third, whereas the percentage of thyroid carcinomas among the patients who underwent surgery doubled.¹⁸⁰ When the results are suggestive of follicular neoplasia, surgery is required because follicular adenoma cannot be differentiated from follicular cancer by cytologic analysis alone. Because the sample obtained is not always representative of the lesion, surgical treatment is indicated for lesions that are highly suspicious for being malignant on clinical grounds. Other uses of aspiration biopsy include presumed lymphoma or invasive anaplastic carcinoma when biopsy might spare the patient an unnecessary neck exploration. Another application of needle aspiration is in the confirmation and treatment of thyroid cysts and autonomous thyroid nodules.¹⁸¹ The latter are best diagnosed by the identification of a mutation in the TSH receptor in material obtained by FNA.^182,183

Evaluation of the Hypothalamic-Pituitary-Thyroid Axis

The development of an RIA for the routine measurement of TSH in serum and the availability of synthetic TRH have placed increased reliance on tests that assess the hypothalamic-pituitary control of thyroid function. These tests allow the diagnosis of mild and subclinical forms of thyroid dysfunction and provide a means of differentiating between primary, pituitary (secondary), and hypothalamic (tertiary) thyroid gland failure.

Thyrotropin

In recent years, dramatic improvements have been made in assays for TSH. The routine measurement of TSH in clinical practice initially used RIA techniques. These first-generation assays had a sensitivity level of 1 mU/L, which did not allow the separation of normal from reduced values. A major problem with early TSH RIAs was cross-reactivity with gonadotropins (luteinizing hormone, follicle-stimulating hormone, and hCG) that share with TSH a common α subunit.¹⁸⁴ Nevertheless, even older RIA methods for measurement of pituitary TSH correlated well with values obtained by bioassay techniques.³¹ Another uncommon source of error is the presence in the serum sample of heterophilic antibodies induced by vaccination with materials contaminated with animal serum or the presence of endogenous TSH antibodies.¹⁸⁵ RIA techniques for the measurement of TSH in dry blood spots on filter paper are used in screening for neonatal hypothyroidism.

Newer techniques have been developed that use multiple antibodies to produce a “sandwich”-type assay in which one antibody (usually directed against the α subunit) serves to anchor the TSH molecule and another (usually monoclonal antibodies directed against the β subunit) is radioiodinated (immunoradiometric assay) or is conjugated with an enzyme (immunoenzymometric) or a chemiluminescent compound (chemiluminescent assay). In these assays, the signal should be directly related to the amount of the ligand present rather than being inversely related as in RIAs that measure the bound tracer. This technique results in decreased background “noise” and greater sensitivity, decreased interference from related compounds, and an expanded useful range. Initial improvements in the TSH assay resulted in assays with a sensitivity limit of 0.1 mU/L, a normal range of approximately 0.5 to 4.5 mU/L, and the ability to distinguish between low and normal TSH values.186,187 Recently, commercial assays have been developed with an even higher sensitivity limit of 0.005 to 0.01 mU/L and a similar normal range but an expanded range between the lower limit of normal and the lower limit of sensitivity.^188,189

The nomenclature for differentiating these various assays has not been standardized, with manufacturers applying various combinations of “high(ly),” “ultra,” and “sensitive.” It has been recommended that the sensitivity limit be used in defining the assays, with the early RIAs that detect values of 1 mU/L or greater being designated first-generation assays, those with a lower sensitivity limit of 0.1 mU/L being designated as second-generation assays, and those with a lower sensitivity limit of 0.01 mU/L or less being designated as third-generation assays. Determination of the appropriate sensitivity level has also been controversial. Some base the definition on the level with a coefficient of variation less than 20%; others define it as the lowest level that can be reliably differentiated from the zero TSH standard. At a minimum, for a TSH assay to be considered “sensitive,” the overlap of TSH values in sera from clinically hyperthyroid and euthyroid individuals should be less than 5% and preferably less than 1%.

In a number of these third-generation assays, TSH that was detected in clinically toxic patients and elevated values that were found in euthyroid subjects were not confirmed when the samples were measured in other assays. In some cases, these discrepant results have been attributed to the presence of antibodies directed against the animal immunoglobulins used in the assay. These immunoglobulins act to bind the anchoring and detecting antibodies and lead to an overestimation of TSH. In some cases, this effect can be blocked by the addition of an excess of nonspecific immunoglobulin of the same species.¹⁹⁰

The availability of random access immunoassay analyzers has revolutionized TSH measurements. These assays are highly reproducible and provide convenience and rapid throughput of large volumes of serum samples.

TSH appears abruptly in the pituitary and serum of the fetus at midgestation and can also be detected in amniotic fluid.191,192 The mean TSH level is higher in cord blood than in maternal blood. A substantial increase, to levels several-fold above the upper range in adults, is observed during the first half-hour of life. Levels decline to near normal adult range by the third day of life. Minimal changes that are reported to occur during adult life and in early adolescence have no significant effect on the overall range of normal. In the absence of pregnancy, no significant gender differences have been observed. Although early studies failed to show diurnal TSH variation, significantly higher values have been recorded during the late evening and early night and are partially inhibited by sleep.¹⁹³ This diurnal rhythm of TSH is superimposed on continuous high-frequency, low-amplitude variations. The nocturnal TSH surge persists in patients with mild primary hypothyroidism^194,195 and is abolished in hypothalamic hypothyroidism^195,196 and in some patients during fasting and with nonthyroidal illness. It is enhanced by oral contraceptives¹⁹⁷ and is abolished by high levels of glucocorticoids.¹⁹³ The presence of seasonal variation has not been a uniform finding, but it is unlikely to affect the clinical interpretation of serum values. Various types of stressful stimuli have no significant effect on the basal serum TSH level, except for a rise during surgical hypothermia in infants but not in adults.¹⁹⁸ Various stimuli that elicit in normal humans a secretory response of some pituitary hormones, such as the administration of insulin, vasopressin, glucagon, bacterial pyrogens, arginine, prostaglandins, and chlorpromazine, have no effect on serum TSH. However, administration of any of a growing list of drugs has been found to alter the basal concentration of serum TSH and/or its response to exogenous TRH (Table 6-10). In the presence of a normally functioning hypothalamic-pituitary system, an inverse correlation is found between the serum concentration of FT₄ and TSH. Changes in the serum concentration of TT₄ as a result of TBG abnormalities or drugs competing with T₄ binding to TBG have no effect on the level of serum TSH. The pituitary is exquisitely sensitive to both minimal decreases and increases in thyroid hormone concentration, with a logarithmic change in TSH levels in response to changes in T₄ (Fig. 6-5). Thus, serum TSH levels should be elevated in patients with primary hypothyroidism and low or undetectable in those with thyrotoxicosis. Indeed, in the absence of hypothalamic pituitary disease, illness, or drugs, TSH is an accurate indicator of thyroid hormone status and the adequacy of thyroid hormone replacement.

Table 6-10

Agents That May Affect TSH Secretion

DT₄, Dextrothyroxine; IL, interleukin; T₃, triiodothyronine; T₄, thyroxine; TRH, thyrotropin-releasing hormone; TRIAC, triiodothyroacetic acid; TSH, thyroid-stimulating hormone.

^*In hyposomatotrophic dwarfs.

FIGURE 6-5 Correlation of the serum thyroid-stimulating hormone (TSH) concentration and the free thyroxine index (FT₄I) in three persons given increasing doses of levothyroxine. Note the logarithmic correlation between TSH and FT₄I and the variable individual requirement of T₄ to normalize the TSH level. Normal ranges are included in the heavily lined box, and those for subjects treated by levothyroxine replacement are in the lightly lined box.

In patients with primary hypothyroidism of whatever cause, levels may reach 1000 mU/L or higher. The magnitude of serum TSH elevation grossly correlates with the severity and, in part, with the duration of thyroid hormone deficiency. TSH concentrations above the upper limit of normal have been observed in the absence of clinical symptoms and signs of hypothyroidism and in the presence of serum T₄ and T₃ levels well within the normal range. This condition is most commonly encountered in patients with incipient hypothyroidism from Hashimoto’s thyroiditis or with limited ability to synthesize thyroid hormone because of prior thyroid surgery, radioiodide treatment, or severe iodine deficiency. No agreement has been reached on whether such patients have subclinical hypothyroidism or a “compensated state” in which euthyroidism is maintained by chronic stimulation of a reduced amount of functioning thyroid tissue through hypersecretion of TSH. Transient hypothyroidism occurs in some infants during the early neonatal period. In two circumstances, the usual inverse relationship between the serum level of TSH and T₄ is not maintained in patients with proven primary hypothyroidism. Treatment with replacement doses of T₄ may normalize or even produce serum levels of thyroid hormone above the normal range before high TSH levels have reached the normal range. This finding is particularly true in patients with severe neonatal or long-standing primary hypothyroidism, who might require 3 to 6 months of hormone replacement before TSH levels are fully suppressed. Conversely, serum TSH concentrations may remain low or normal for up to 5 weeks after withdrawal of thyroid hormone replacement when serum levels of T₄ and T₃ have already declined to values well below the lower range of normal. Causes of discrepancies between TSH and FT₄ and FT₃ levels are listed in Table 6-11.

At this time, it is uncertain what TSH level is appropriate for suppressive thyroid hormone therapy. The frequency with which patients have subnormal but detectable TSH values depends on both the population studied and the sensitivity of the assay (Fig. 6-6). When an assay is used with a sensitivity limit of 0.1 mU/L, 3% to 4% of hospitalized patients have been noted to have subnormal TSH. When patients with undetectable TSH in such an assay were reevaluated in an assay with a sensitivity limit of 0.005 mU/L, 3 of 77 (4%) with thyrotoxicosis and 32 of 37 (86%) with nonthyroidal illness or taking drugs were found to have a subnormal, but detectable, TSH level.¹⁸⁹ Thus, the more sensitive the assay, the more likely it is that patients with clinical thyrotoxicosis will have undetectable serum TSH, whereas those with illness will have a subnormal but detectable level. However, with progressively more sensitive assays, the likelihood that a clinically toxic patient will have detectable TSH will increase, and if patients who are receiving suppressive therapy are treated until the TSH is undetectable, they are more likely to have symptoms of thyrotoxicosis.

FIGURE 6-6 The effect of serum thyroid-stimulating hormone (TSH) assay sensitivity on the discrimination of euthyroid subjects (Euth) from those with thyrotoxicosis (Toxic). (Data from Spencer C: Clinical Diagnostics. Rochester, NY, Eastman Kodak, 1992.)

Persistent absence of a reverse correlation between the serum thyroid hormone and the TSH concentration has a very different connotation. A low serum level of thyroid hormone without clear elevation in the serum TSH concentration is suggestive of trophoprivic hypothyroidism (central or secondary hypothyroidism), especially when associated with obvious clinical stigmata of hypothyroidism.

Resistance to TSH is an inherited syndrome of variable thyroid sensitivity to a biologically active TSH molecule. The clinical features of this condition are (1) elevated serum TSH values with normal biological activity when tested in vitro, (2) absence of goiter but the presence of a normal or hypoplastic gland, and (3) normal (“compensated”) or low (“not compensated”) TSH levels depending on the degree of TSH insensitivity. Therefore, subjects with resistance to TSH can present as euthyroid hyperthyrotropinemia or with severe primary hypothyroidism. The index case of the first reported family with resistance to TSH due to a mutation in the thyroid-stimulating hormone receptor (TSHR) presented with a blood TSH on neonatal screening of 103 mU/L (normal, <20). Repeat measurement on the sixteenth day of life was 47 mU/L with a normal T₄; a radioiodide scan reveals a thyroid gland in the normal location and of normal size. The mother was found to be heterozygous for an abnormal TSHR, P162A, and the father was found to be heterozygous for another abnormal TSHR, I167N. All the children were compound heterozygous for the P162A and I167N and therefore had significant elevations of serum TSH.¹⁹⁹ In other cases of resistance to TSH, the TSH receptor gene had a normal sequence, and the molecular basis for the phenotype remains unknown.

In some cases, a mild elevation in the serum TSH level measured by RIA is probably due to the presence of immunoreactive TSH with reduced biological activity. Distinction between pituitary and hypothalamic hypothyroidism may be made on the basis of the TSH response to the administration of TRH, although the responses can be very variable (Table 6-12, and see below).

Table 6-12

Differential Diagnosis of Euthyroid Hyperthyrotropinemia

AITD, Autoimmune thyroid disease; Chol, cholesterol; THs, thyroid hormones T₄ and T₃; RTSH, resistance to TSH; TRH, thyrotropin-releasing hormone; TSH, thyroid-stimulating hormone; TSHR, thyroid-stimulating hormone receptor.

In another group of pathologic conditions, serum TSH levels might not be suppressed despite a clear elevation of serum free thyroid hormone levels. Because such a finding is incompatible with a normal thyroregulatory control mechanism of the pituitary, which is preserved in the more common forms of thyrotoxicosis, it has been termed inappropriate secretion of TSH. It implicitly suggests defective feedback regulation of TSH. When associated with the clinical and metabolic changes of thyrotoxicosis, it is usually due to TSH-secreting pituitary adenoma and rarely to partial resistance to thyroid hormone. The existence of hypothalamic hyperthyroidism can be questioned. Precise diagnosis requires additional studies, including radiologic examination of the pituitary gland and a TRH test. In addition, the presence of high circulating levels of the common α subunit (α-SU) of the pituitary glycoprotein hormones, with a subsequent disproportionately high α-SU/TSH molar ratio in serum, is characteristic, if not pathognomonic, of TSH-secreting pituitary tumors. Normal and occasionally high serum TSH levels associated with a clear elevation in serum FT₄ and FT₃ but no clear clinical evidence of hypothyroidism or symptoms and signs suggestive of both thyroid hormone deficiency and excess are typical of resistance to thyroid hormone.

Although TSH has been implicated in the pathogenesis of simple, nontoxic goiter, unless hypothyroidism supervenes or iodide deficiency is very severe, TSH levels are characteristically normal. Elevated TSH levels may occur in the presence of normal thyroid hormone levels and apparent euthyroidism in nonthyroidal diseases and with primary adrenal failure. A more common occurrence in severe acute and chronic illnesses is a normal or low serum TSH concentration despite low levels of T₃ and even low T₄ levels. TSH values may be transiently elevated during the recovery phase. Various hypotheses to explain these anomalous findings have been proposed, but a satisfactory explanation is not at hand.

A specific RIA for the β subunits of human TSH is also available but has not found clinical application. Measurement of TSH usually consists of the quantitation of the immunoreactive moiety. However, under certain physiologic and pathologic states, there can be a change in the ratio of immunoreactive and bioreactive TSH. For example, the nocturnal rise in TSH is due to less bioreactive TSH because of different glycosylation. In addition, in states of hypothalamic dysfunction, there is also a decrease in the bioreactive-to-immunoreactive ratio of TSH, usually due to an alteration in the degree of glycosylation of the TSH molecule. Bioassay of TSH is performed by using JP26₂₆ cells. This is a subclone of JP26, a line of Chinese hamster ovary cells that are stably transfected with a human TSH receptor cDNA. Standard bovine TSH or samples for TSH measurement are added with Rolipram, a cyclic adenosine monophosphate (cAMP) phosphodiesterase inhibitor. cAMP is measured in the dried cell extract by RIA.

Thyrotropin-Releasing Hormone

The hypothalamic tripeptide TRH (protirelin) plays a central role in the regulation of pituitary TSH secretion. Several methods have been used for quantitation of TRH, but for many reasons measurement in humans has failed to provide information of diagnostic value. These reasons include high dilution of TRH by the time that it reaches the systemic circulation, rapid enzymatic degradation, and ubiquitous tissue distribution. Mean serum TSH levels of 5 and 6 pg/mL have been reported. It is uncertain whether measurements carried out in urine truly represent TRH.²⁰⁰

The TRH test measures the increase in pituitary TSH in serum in response to the administration of synthetic TRH. The magnitude of the TSH response to TRH is modulated by the thyrotroph response to active thyroid hormone and is thus almost always proportional to the concentration of free thyroid hormone in serum. The response is exquisitely sensitive to minor changes in the level of circulating thyroid hormones, which might not be detected by direct measurement. A direct correlation between basal serum TSH values and the maximal response to TRH has been observed even in the absence of thyroid hormone abnormalities, which suggests that the euthyroid state might be associated with a fine modulation of pituitary sensitivity to TRH.²⁰⁰

TRH normally stimulates pituitary prolactin secretion, and under certain pathologic conditions, the release of growth hormone and adrenocorticotropic hormone. Accordingly, the test has been used for the assessment of a variety of endocrine functions, some unrelated to the thyroid. In clinical practice, the TRH test is used mainly (1) to assess the functional integrity of pituitary thyrotrophs and thus to aid in differentiating hypothyroidism caused by intrinsic pituitary disease from hypothalamic dysfunction, (2) in the diagnosis of mild thyrotoxicosis when results of other tests are equivocal, and (3) in the differential diagnosis of inappropriate TSH secretion, in particular, when a TSH-secreting adenoma is suspected.

TRH is effective when given intravenously as a bolus or by infusion,²⁰¹ intramuscularly,²⁰² or orally in single or repeated doses. Doses as small as 6 µg can elicit a significant TSH response, and a linear correlation exists between the incremental changes in serum TSH concentration and the logarithm of the TRH dose administered. The standard test uses a single TRH dose of 200 or 400 µg/1.73 m² body surface area, given by rapid intravenous injection. Serum is collected before and at 15 minutes and then at 30 minute intervals over a period of 120 to 180 minutes, although many clinicians choose to obtain a single postinjection sample at 15, 20, or 30 minutes. Normal individuals have a prompt increase in serum TSH, with a peak level at 15 to 40 minutes that is on the average 16 mU/L, or fivefold the basal level. The decline is more gradual, with a return of serum TSH to the preinjection level by 3 to 4 hours. Results can be expressed in terms of the peak level of TSH achieved, the maximal increment above the basal level, the peak TSH value expressed as a percentage of the basal value, or the integrated area of the TSH response curve. Determination of TSH before and 30 minutes after injection of TRH provides information concerning the presence or absence of TSH responsiveness but cannot detect delayed or prolonged responses.

At the time of this writing, TRH was not available in the United States. It can, however, be obtained for use in the United States after submission of an Investigational New Drug application from the Food and Drug Administration (FDA). It is manufactured by Ferring Arzneimittel, GmbH (Wittland 11; 24109 Kiel, Germany; telephone number: 0049 431 58520; fax number: 0049 431 585235) or the distributor UNIPHARMA SA (Via Pian Scairolo, 6; 6917 Barbengo, Switzerland).

The stimulatory effect of TRH is specific for pituitary TSH, its free α and β subunits, and prolactin. Under normal circumstances, no significant changes are observed in the serum levels of other pituitary hormones ²⁰³ or potential thyroid stimulators.²⁰⁴ Responsiveness is present at birth,²⁰⁵ is greater in women than in men, particularly in the follicular phase of the menstrual cycle, and may be blunted in older men, but this finding is not consistent. On average, the magnitude of the response is greater at 11:00 pm than at 11:00 am, in accordance with the diurnal pattern of the basal TSH level, which correlates with its response to TRH. Repetitive administration of TRH to the same subject at daily intervals causes gradual blunting of the TSH response, presumably because of the increase in thyroid hormone concentration and also in part because of TSH “exhaustion.” However, more than 1 hour must elapse between the increase in thyroid hormone concentration and TRH administration for inhibition of the TSH response to occur. A number of drugs and nonendocrine diseases may affect the magnitude of the response to varying extent.

TRH-induced secretion of TSH is followed by the release of thyroid hormone that can be detected by direct measurement of serum TT₄ and TT₃ concentrations. Peak levels are normally reached approximately 4 hours after the administration of TRH and are accompanied by an increase in serum Tg concentration. The incremental rise in serum TT₃ is relatively greater, and the peak is on average 50% above the basal level. Measurement of changes in serum thyroid hormone concentration after the administration of TRH has been proposed as an adjunctive test and is useful for evaluation of the integrity of the thyroid gland or the bioactivity of endogenous TSH.²⁰⁶ The increase in RAIU is minimal and occurs only with high doses of TRH given orally.

Side effects from the intravenous administration of TRH, in decreasing order of frequency, include nausea, flushing or a sensation of warmth, desire to micturate, peculiar taste, light-headedness or headache, dry mouth, urge to defecate, and chest tightness. They are usually mild, begin within a minute after the injection of TRH, and last for a few seconds to several minutes. A transient rise in blood pressure has been observed on occasion, but no other changes are seen in vital signs, urine analysis, blood count, or routine blood chemistry tests.²⁰⁷ The occurrence of circulatory collapse is exceedingly rare. There have been several case reports of pituitary apoplexy in patients who received TRH testing with or without gonadotropin-releasing hormone. In many of these cases, the patient had a large pituitary tumor, which might predispose to this complication.

The test provides a means to distinguish between secondary (pituitary) and tertiary (hypothalamic) hypothyroidism (Fig. 6-7). Although the diagnosis of primary hypothyroidism can be easily confirmed by the presence of elevated basal serum TSH levels, secondary and tertiary hypothyroidism are typically associated with TSH levels that are low or normal. On occasion, the serum TSH concentration might be slightly elevated because of the secretion of biologically less potent molecules, but it remains inappropriately low for the degree of thyroid hormone deficiency. Differentiation between secondary and tertiary hypothyroidism cannot be made with certainty without the TRH test. In a case report of a family with congenital hypothyroidism secondary to hemizygous mutations in the TSH receptor gene, response of TSH to TRH was normal.²⁰⁸ A TSH response is suggestive of a hypothalamic disorder, and failure to respond is compatible with intrinsic pituitary dysfunction. Furthermore, the typical TSH response curve in hypothalamic hypothyroidism shows a delayed peak with a prolonged elevation in serum TSH before return to the basal value (see Fig. 6-7). The finding of a lack of TSH response in association with normal prolactin stimulation may be due to isolated pituitary TSH deficiency. Caution should be exercised in the interpretation of test results after withdrawal of thyroid hormone replacement or after treatment of thyrotoxicosis because despite a low serum thyroid hormone concentration, TSH may remain low and may not respond to TRH for several weeks. In the most common forms of thyrotoxicosis, the mechanism of feedback regulation of TSH secretion is intact but is appropriately suppressed by the excessive amounts of thyroid hormone. Thus, both the basal TSH level and its response to TRH are suppressed unless thyrotoxicosis is TSH induced. With the development of more sensitive TSH assays, the TRH test generally is not needed in the evaluation of a thyrotoxic patient with undetectable TSH. The differential diagnosis of conditions leading to inappropriate secretion of TSH may be aided by the TRH test result. Elevated basal TSH values that do not respond to TRH by a further increase are typical of TSH-secreting pituitary adenomas. In contrast, patients with high thyroid hormone levels but detectable serum TSH resulting from resistance to thyroid hormone have a normal or exaggerated TSH response to TRH that in most instances is suppressed by supraphysiologic doses of thyroid hormone. Because of the exquisite sensitivity of the pituitary gland to feedback regulation by thyroid hormone, small changes in the latter profoundly affect the response of TSH to TRH. Thus, patients with non–TSH-induced thyrotoxicosis of the mildest degree have a reduced TSH response to TRH, whereas those with primary hypothyroidism exhibit an accentuated response that is prolonged (see Fig. 6-7). These changes can occur in the absence of clinical or other laboratory evidence of thyroid dysfunction.

FIGURE 6-7 Typical serum thyroid-stimulating hormone (TSH) responses to the administration of a single intravenous bolus of thyrotropin-releasing hormone at time 0 in various conditions. The normal response is represented by the shaded area. Data used for this figure are the average of several studies.

The TSH response to TRH is subnormal or absent in one third of apparently euthyroid patients with autoimmune thyroid disease, and even members of their family might not respond to TRH. Most, but not all, patients with a reduced TSH response to TRH will also show thyroid activity that is nonsuppressible by thyroid hormone. A common dissociation between these two tests is typified by a normal TSH response to TRH in a nonsuppressible patient. This finding is not surprising because patients with nonsuppressible thyroid glands often have limited capacity to synthesize and secrete thyroid hormone as a result of prior therapy or partial destruction of their glands by the disease process. Clinically, euthyroid patients who do not respond to TRH admittedly have a slight excess of thyroid hormone. It is less easy to reconcile the rare occurrence of TRH unresponsiveness in a patient whose TSH is suppressible by exogenous thyroid hormone. It should be remembered, however, that a suppressed pituitary can take a variable amount of time to recover, a phenomenon that might be the basis of such discrepancies. Despite discrepancies between the results of the TRH and T₃ suppression tests, use of the former is much preferred, particularly in elderly patients, in whom administration of T₃ can produce untoward effects.

Other Tests of TSH Reserve

It has been reasoned that by virtue of different mechanisms of action, testing the TSH response by means other than TRH can provide information of diagnostic value that is not obtainable from stimulation and suppression of the pituitary by TRH and thyroid hormone, respectively. Trials using drugs such as metoclopramide and l-dopa have been carried out but so far have provided only limited additional information and so have not found a place in clinical practice. These tests have limited application in the study of patients with inappropriate secretion of TSH, in whom the distinction of autonomous secretion of TSH as compared with selective unresponsiveness to thyroid hormone inhibition is of diagnostic value.

Other tests indirectly measure pituitary TSH reserve during the rebound period after suppression of thyroid hormone synthesis or pituitary TSH secretion. Assessment of thyroid gland activity after withdrawal of antithyroid drugs or T₃ replacement has been proposed.209,210

Thyrotropin Stimulation Test

The TSH stimulation test measures the ability of thyroid tissue to respond to exogenous TSH by an increase in iodide accumulation and hormone release and for the identification of residual thyroid cancer tissue. Formerly used to differentiate hypothyroidism caused by thyroid gland failure from that caused by TSH deficiency, the test was recently used in conjunction with a scintiscan to localize areas of suppressed thyroid tissue. Formerly, it required the intramuscular administration of one or three 5 to 10 U doses of bovine TSH. The test could cause discomfort, and even serious reactions to the heterologous TSH, and therefore is no longer used.

Recombinant human TSH (r-hTSH), Thyrogen (Genzyme Transgenics Corp., Cambridge, MA), was approved by the Food and Drug Administration over 6 years ago for the management of patients with thyroid cancer. Many studies have subsequently demonstrated the utility of r-hTSH to stimulate thyroid remnants, but few have evaluated its effect on intact thyroids. A single dose of 0.1 mg r-hTSH (compared with 0.9 mg given to thyroid cancer patients) was a potent stimulus for the release of T₄, T₃, and Tg in normal volunteers.²¹¹ A very low dose of r-hTSH (0.01 mg) has been found to stimulate ¹³¹I uptake in multinodular goiters, but no data are available on whether this very low dose is capable of stimulating TSH or Tg release from the thyroid gland. It has also been suggested that dosing of r-hTSH should be based on body surface area (BSA) as an inverse relationship between BSA and serum peak TSH after r-hTSH administration has been reported.²¹²

Thyroid Suppression Test

Maintenance of thyroid gland activity that is independent of TSH can be demonstrated by the thyroid suppression test. Under normal conditions, administration of thyroid hormone in quantities sufficient to satisfy the body requirement suppresses endogenous TSH and results in a reduction in thyroid hormone synthesis and secretion. Because thyrotoxicosis resulting from excessive secretion of hormone by the thyroid gland implies that the feedback control mechanism is not operative or has been perturbed, it is easy to understand why, under such circumstances, the supply of exogenous hormone would also be ineffective in suppressing thyroid gland activity. The test has very limited application today, although it might be of value in patients who are euthyroid or only mildly thyrotoxic but are suspected of having abnormal thyroid gland stimulation or autonomy, particularly for confirmation of the diagnosis of resistance to thyroid hormone.

Usually, the test is carried out with 100 µg of liothyronine (L-T₃) given daily in two divided doses over a period of 7 to 10 days. A 24 hour RAIU is determined before and during the last 2 days of T₃ administration.²¹³ Normal individuals show suppression of RAIU by at least 50% when compared with the value before liothyronine treatment. No change or a lesser reduction is typical of not only Graves’ disease but also other forms of endogenous thyrotoxicosis, including toxic adenoma, functioning carcinoma, and thyrotoxicosis caused by trophoblastic disease. The presence of nonsuppressibility indicates thyroid gland activity independent of TSH but not necessarily thyrotoxicosis. Euthyroid patients with autonomous thyroid function have a normal TSH response to TRH before the administration of liothyronine. However, inhibition of TSH secretion by exogenous T₃ does not suppress the autonomous activity of the thyroid gland. This discrepancy is the most commonly encountered difference between the results of the two related tests. When the T₃ suppression test is used in conjunction with a scintiscan, localized areas of autonomous function can be identified. The test can be carried out without the administration of radioisotopes by measuring serum T₄ before and 2 weeks after the ingestion of liothyronine. Although total suppression of T₄ secretion never occurs, even after prolonged treatment with liothyronine, reduction by at least 50% is normal.²¹⁴

Variants of the test have been proposed to reduce the potential risks of liothyronine administration in elderly patients and in those with angina pectoris or congestive heart failure. However, with the availability of sensitive TSH assays, thyroid suppression tests are no longer indicated.