Measurement of Total Thyroid Hormone Concentration in Serum

Iodometry. Iodine constitutes an integral part of the thyroid hormone molecule. It is thus not surprising that determination of iodine content in serum was the first method suggested almost six decades ago for the identification and quantitation of thyroid hormone.²⁷ Measurement of the Protein-Bound Iodine (PBI) 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 the serum proteins,²⁸ 90% of which is T4. The normal range was 4 - 8 µg I/dl 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 the butanol extractable iodine (BEI) and T4I by column techniques. All such chemical methods for the measurement of thyroid hormone in serum have been replaced by the ligand assays which are devoid of interference by even large quantities of nonhormonal iodine-containing substances.

Radioimmunoassays. Concentrations of thyroid hormones in serum can be measured by radioimmunoassays (RIA). The principle of these assays is the competition of a hormone (H), being measured, with the same isotopically labeled compound (H*) for binding to a specific class of IgG molecules present in the antiserum [antibody (Ab)]. H is the ligand and the Ab is either a polyclonal antiserum to H or a monoclonal IgG. The reaction obeys the law of mass action. Thus, at equilibrium, the amount of H* bound to Ab to form the complex Ab-H* is inversely proportional to the concentration of H, forming the complex Ab-H, provided the amounts of Ab and H* are kept constant.

AbH* + [H] AbH + H*

The radioisotope content in Ab-H* or in the unbound (free) H* is determined after their separation by precipitation of the antibody-ligand complex or adsorption of the free ligand. Some RIAs are carried out with the Ab fixed to a solid support, reacting with H and H* in solution. Increments of known amounts of H are added to a series of reactions to construct a standard curve that describes the curvilinear stoichiometric relationship between Ab-H* and H. It can be converted to a straight line by a number of mathematical transformations, such as the logit-log plot. Blank reactions contain H* but not specific Ab or, a large excess of H in a full reaction.²⁹ The sensitivity of the assay is dependent upon the affinity of the Ab and specific activity of H*. Under optimal conditions, as little as 1 pg of H can be measured.

In assays for thyroid hormones, the hormone needs to be liberated from serum binding proteins, mainly TBG. Methods to achieve this include extraction, competitive displacement of the hormone being measured, or inactivation of thyroxine-binding globulin (TBG).^31-34 Rarely, some patients develop circulating antibodies against thyronines that interfere with the RIA carried out on unextracted serum samples. Depending on the method used for the separation of bound from free ligand, values obtained may be either spuriously low or high in the presence of such antibodies.^38,39

A wide choice of commercial kits is available for most RIA procedures, making these assays accessible to all medical centers. RIAs have been adapted for the measurement of T4 in small samples of dried blood spots on filter paper and are used in screening for neonatal hypothyroidism.⁴⁰

Non-radioactive Methods. More recently, assays have been developed that are based on the principle of the radioligand assay but do not use radioactive material. These assays, which use ligand conjugated to an enzyme have largely replaced RIAs. The enzyme-linked ligand competes with the ligand being measured for the same binding sites on the antibody. Quantitation is carried out by spectrophotometry of the color reaction developed after the addition of the enzyme substrate.⁴² Both homogeneous [enzyme-multiplied immunoassay technique (EMIT)] and heterogeneous [enzyme-linked immunosorbent assay (ELISA)] assays for T4 have been developed.^43-45 In the homogeneous assays, no separation step is required, thus providing easy automation.⁴³ In one such assay, T4 is linked to malate dehydrogenase, inhibiting the enzyme activity. The enzyme is activated when the T4-enzyme conjugate is bound to T4-specific antibody. Active T4 conjugates to other enzymes, such as peroxidase⁴⁴ and alkaline phosphatase,⁴⁵ have also been developed. The assay has been adapted for the measurement of T4 in dried blood samples used in mass screening programs for neonatal hypothyroidism.⁴⁵ Other non-radioisotope immunoassays use fluorescence excitation for detection of the labeled ligand, a technique which is finding increasing application. Such assay methods utilize a variety of chemiluminescent molecules such as 1,2-dioxetanes, luminol and derivatives, acridinium esters, oxalate esters and firefly luciferins, as well as many sensitizers and fluorescent enhancers.^45a One such assay which employes T4 conjugated to ß-galactosidase and fluoresence measurements of the hydrolytic product of 4-methyl-umbelliferyl-ßD-galactopyranoside has been adapted for use in a microanalytical system requiring only 10µl of serum.^45b

Serum Total Thyroxine (TT4). The usual concentration of TT4 in adults ranges from 5 to 12 µg/dl (64 - 154 nmol/L). When concentrations are below or above this range in the absence of thyroid dysfunction, they are usually the result of an abnormal level of serum TBG. The hyperestrogenic state of pregnancy and administration of estrogen-containing compounds are the most common causes of a significant elevation of serum TT4 levels in euthyroid persons. Less commonly, TBG excess is inherited.⁵⁰ Serum TT4 is virtually undetectable in the fetus until midgestation. Thereafter, it rapidly increases, reaching high normal adult levels during the last trimester. A further acute but transient rise occurs within hours after delivery.⁵¹ Values remain above the adult range until 6 years of age, but subsequent age related changes are minimal so that in clinical practice, the same normal range of TT4 applies to both sexes and all ages.

Small seasonal variations and changes related to high altitude, cold, and heat have been described. Rhythmic variations in serum TT4 concentration are of two types: variations related to postural changes in serum protein concentration⁵⁶ and true circadian variation.³¹ Postural changes in protein concentration do not alter the free T4 (FT4) concentration.

Although levels of serum TT4 below the normal range are usually associated with hypothyroidism, and above this range with thyrotoxicosis, it must be remembered that the TT4 level may not always correspond to the FT4 concentration which represents the metabolically active fraction (see below). The TT4 concentration in serum may be altered by independent mechanisms: (1) an increase or decrease in the supply of T4 , as seen in most cases of thyrotoxicosis and hypothyroidism, respectively; (2) changes due solely to alterations in T4 binding to serum proteins; and (3) compensatory changes in serum TT4 concentration due to high or low serum levels of T3. Conditions associated with changes in serum TT4 and their relationship to the metabolic status of the patient are listed in Table 6-4.

Serum TT4 levels are low in conditions associated with decreased TBG concentration, the presence of abnormal TBG's with reduced binding affinity (see Chapter 16) or when the available T4-binding sites on TBG are partially saturated by competing drugs present in blood in high concentrations (see Table 5-2). Conversely, TT4 levels are high when the serum TBG concentration is high. The person remains euthyroid provided the feedback regulation of the thyroid gland is intact.

Although changes in transthyretin (TTR) concentration rarely give rise to significant alterations in TT4 concentration,⁵⁷ the presence of a variant serum albumin with high affinity for T4 ^58,59 or antibodies against T4 ^38,39 produce apparent elevations in the measured TT4 concentration, whereas the metabolic status remain normal. The variant albumin is inherited as an autosomal dominant trait termed familial dysalbuminemic hyperthyroxinemia (FDH) (see Chapter 16).

Another possible cause of discrepancy between the observed serum TT4 concentration and the metabolic status of the patient is divergent changes in the serum TT3 and TT4 concentrations with alterations in the serum T3/T4 ratio. The most common situation is that of 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.⁶⁰ This finding indicates that in T3 thyrotoxicosis the hormone is predominantly secreted as such rather than arising from 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. A moderate hypersecretion of T3 can be associated with euthyroidism and a low serum TT4 concentration. This circumstance, 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 for thyrotoxicosis, in some cases of chronic thyroiditis, or during iodine deprivation.^61,62 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 15-50% of patients treated with exogenous T4.⁶³ Because of the relatively slow rate of metabolism and 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.⁶⁴

Serum Total Triiodothyronine (TT3). Normal serum TT3 concentrations in the adult are 80-190 ng/dl (1.2 - 2.9 nmol/L). While sex differences are small, those with age are more dramatic. In contrast to serum TT4, TT3 concentration at birth is low, about one-half the normal adult level. It rises within 24 hours to about double the normal adult value followed by a rapid decrease over the subsequent 24 hours to a level in the upper adult range, which persists for the first year of life.⁵¹ A decline in the mean TT3 level has been observed in old age, although not in healthy subjects.^52,53 so that a fall in TT3 may refelct the prevalence of nonthyroidal illness rather than to age alone.⁶⁷ Although a positive correlation between serum TT3 level and body weight has been observed, it may be related to overeating.⁶⁸ Rapid and profound reductions in serum TT3 level can be produced within 24-48 hours of total calorie or only carbohydrate deprivation.^69-71

Most conditions causing 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.⁷² Accordingly, measurement of the serum TT3 level is a more sensitive test for the diagnosis of hyperthyroidism, and that of TT4 more useful in the diagnosis of hypothyroidism.

There are circumstances in which changes in the serum TT3 and TT4 concentrations are either disproportionate or in opposite direction (Table 6-5). These include the syndrome of thyrotoxicosis with normal TT4 and FT4 levels (T3 thyrotoxicosis). In some patients, treatment of thyrotoxicosis with antithyroid drugs may normalize the serum TT4 but not TT3 level, producing a high TT3/TT4 ratio. In areas of limited iodine supply ⁶² and in patients with limited thyroidal ability to process iodide,⁶¹ 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.

The most common cause of discordant serum concentrations of TT3 and TT4 is a selective decrease of serum TT3 due to 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 (see Chapter 5). In these conditions, the serum TT3 level is often lower than that commonly found in patients with frank primary hypothyroidism. Yet, these persons do not present clear clinical evidence of hypometabolism. In some individuals, decreased T4 to T3 conversion in the pituitary gland⁷⁵ or in peripheral tissues⁷⁶ is thought to be an inherited condition.

A variety of drugs may also produce changes in the serum TT3 concentration without apparent metabolic consequences (see Chapter 6). Drugs that compete with hormone binding to serum proteins decrease serum TT3 levels, generally without affecting the free T3 concentration (Table 5-5). Some drugs, such as glucocorticoids,⁷⁷ depress the serum TT3 concentration by interfering with the peripheral conversion of T4 to T3. Others, such as phenobarbital,⁷⁸ depress the serum TT3 concentration by stimulating the rate of intracellular hormone degradation. The majority have multiple effects. These effects are combinations of those described above, as well as inhibition of the hypothalamic-pituitary axis or thyroidal hormonogenesis.⁷⁹

Changes in serum TBG concentration have an effect on the serum TT3 concentration similar to that on TT4 (see Chapter 16). The presence of endogenous antibodies to T3 may result in apparent elevation of the serum TT3 but as in the case of high TBG, it does not cause hypermetabolism.³⁸

Administration of commonly used replacement doses of T3, usually in the order of 75 µg/day or 1 µg/kg body weight per day,⁸⁰ results in serum TT3 levels in the thyrotoxic range. Furthermore, because of the rapid gastrointestinal absorption and relatively fast degradation rate, the serum level varies considerably according to the time of sampling in relation to hormone ingestion.⁶⁴

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

Because 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, the estimation of the latter has proved useful in the correct interpretation of values obtained from the 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 concentration due to alterations of binding proteins in euthyroid patients from those due to abnormalities in thyroid gland activity giving 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 are often used as absorbents, the test became known as the resin T3 or T4 uptake test (T3U or T4U), describing 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 not tightly bound to TBG.⁸³ Labeled T3 is usually used because of its less firm yet preferential binding to TBG. Depending upon the method, typical normal results for T3U are 25-35% or 45-55%. 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 0.85-1.15.

The uptake of the tracer by the absorbent is inversely proportional to the amount of unsaturated binding sites (unoccupied by endogenous thyroid hormone) in serum TBG. Thus, the 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, the 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. Since 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 both serum TT4 concentration and the T3U test indicate hyperthyroidism and hypothyroidism, respectively, whereas discrepant changes in serum TT4 and T3U suggest abnormalities in TBG binding. However, abnormalities in hormone and TBG concentrations may coexist in the same patient. For example, a hypothyroid patient with a low TBG level will typically show a low TT4 level and normal T3U result (Figure 6-4). Several nonhormonal compounds, due to structural similarities, compete with thyroid hormone for its binding site on TBG. Some are used as pharmacologic agents and may thus alter the in vitro uptake test as well as the total thyroid hormone concentration in serum. A list is provided in Table 5-2.

Figure 6-4. Graphic representation of the relationship between the serum total T4 concentration, the RT3U test, and the free T4 (FT4) concentration in various metabolic states and in association with changes in 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 and black dots, 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. (From S. Refetoff, Endocrinology, L.J. DeGroot (ed). 1979, Grune & Straton Inc.)

TBG and TTR Measurements.

The concentrations of TBG and TTR in serum can be either estimated by measurement of their total T4-binding capacity at saturation or more usually measured directly by immunologic techniques.^87,88

TBG concentration in serum can be determined by RIA,⁸⁸ and both TBG and TTR can be measured by Laurell's rocket immunoelectrophoresis,^89,90 by radial immunodiffusion,⁹¹ or by enzyme immunoassay;⁸⁷ commercial methods are available. The true mean value for TBG is 1.6 mg/dl (260 nmol/L), with a range of 1.1 - 2.2 mg/dl(180 - 350 nmol/L) serum. In adults, the normal range for TTR is 16 - 30 mg/dl (2.7 - 5.0 µmol/L). The concentrations of TBG and TTR in serum vary with age, sex, pregnancy, and posture. Determination of the concentration of these proteins in serum is particularly helpful in 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

A minute amount of thyroid hormone circulates in the blood in a free form, not bound to serum proteins. It is in reversible equilibrium with the bound hormone and represents the diffusible fraction of the hormone capable of traversing cellular membranes to exert its effects on body tissues.⁹⁴ Although changes in serum hormone-binding proteins affect both the total hormone concentration and the corresponding fraction circulating free, in the euthyroid person the absolute concentration of free hormone remains constant and correlates with the tissue hormone level and its biologic effect. Information concerning this value is probably the most important parameter in the evaluation of thyroid function as it relates to the metabolic status of the patient.

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,⁹⁷ provided the patient is in a steady state (see Fig. 5-4). Notably, free T4 (FT4) concentration may be normal or even low in patients with T3 thyrotoxicosis and in those ingesting pharmacologic doses of T3. On occasion, the concentration of FT4 may be outside the normal range in the absence of an apparent abnormality in the thyroid hormone-dependent metabolic status. This is frequently observed in severe nonthyroidal illness during which both high and low values have been reported.^98-100 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.¹⁰¹ This decrease is characteristic of all conditions associated with depressed serum TT3 concentrations due to a diminished conversion of T4 to T3 in peripheral tissues (see Chapter 5). 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 resistance to thyroid hormone (see Chapter 16). The FT3 concentration is usually normal or even high in hypothyroid persons living in areas of severe endemic iodine deficiency. Their FT4 levels are, however, normal or low.⁶²

Direct Measurement of Free T4 and Free T3. Direct measurements of the absolute FT4 and FT3 concentrations are technically difficult and have, until recently, been limited to research assays. In order to minimize perturbations of the relationship between the free and bound hormone, these must be separated by ultrafiltration or by dialysis involving minimal dilution and little alteration of the pH or electrolyte composition. The separated free hormone is then measured directly by radioimmunoassay or chromatography.^97,97a 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 may not fully reflect the free concentration in vivo.

Isotopic Equilibrium Dialysis. This method has been the "gold standard" for the estimation of the FT4 or FT3 concentration for almost 30 years. It is based on the determination of proportion of T4 or T3 that is unbound, or free, and is thus able to diffuse through a dialysis membrane, i.e., the dialyzable fraction (DF). To carry out the test, a sample of serum is incubated with a tracer 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 DF 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 purification⁹⁸ and by various techniques of precipitation of the dialyzed hormone.102 FT4 and FT3 levels can be measured simultaneously by addition to the sample of T4 and T3 labeled with two different radioiodine isotopes.103 Ultrafiltration is a modification of the dialysis technique.⁹⁸ Results are expressed as the fraction (DFT4 or DFT3) or percent (%FT4 or %FT3) of the respective hormones which 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 DF. Typical normal values for FT4 in the adult range from 1.0 to 3.0 ng/dl (13 - 39 pmol/L) and for FT3 from 0.25 to 0.65 ng/dl (3.8 - 10 nmol/L).

Results by these techniques are generally comparable to those determined with the 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 non-thyroidal illness.^{104, 104a,104b} The measured DF 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.¹⁰⁵ The calculated value is dependent on an accurate measurement of total T4 or T3 and may be incorrect in patients with T4 or T3 autoantibodies. Some of these problems, particularly those arising from dilution, may be superceded by commercially available dialysis methods or ultrafiltration methods of free from bound hormone which do not necessitate serum dilution.

Index Methods. As the determination of free hormone by equilibrium dialysis is cumbersome and technically demanding, many clinical laboratories have used a method by which a free T4 index (FT4I) or free T3 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 below). While 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 the presence of circulating inhibitors of protein binding , and their reliability has been questioned in patients with non-thyroidal illness.

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 is true provided results of the in vitro uptake test (T3U or T4U) are expressed as the thyroid hormone binding ratio (THBR), determined by dividing the tracer counts bound to the solid matrix by counts bound to serum proteins.¹⁰⁶ Values are corrected for assay variations using appropriate serum standards and are expressed as the ratio of a normal reference pool.^106,107 The normal range is slightly narrower than the corresponding TT4 in healthy euthyroid patients with a normal TBG concentration. It is 6.0 - 10.5 µg/dl or 77 - 135 nmol/l when calculated from TT4 values measured by RIA. In thyrotoxicosis, FT4I is high and in hypothyroidism it is low irrespective of the TBG concentration. Euthyroid patients with TT4 values outside the normal range as a result of TBG abnormalities have a normal FT4I.⁸³ Lack of correlation between the FT4I and the metabolic status of the patient has been observed under the same circumstances as those described for similar discrepancies when the FT4 concentration was measured by dialysis.

Methods for the estimation of the FT3I are also available¹⁰³ 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 FT4 and FT3 Based on TBG Measurements. Since most T4 and T3 in serum are bound to TBG, their free concentration can be calculated from their binding affinity constants to TBG and molar concentrations of hormones and TBG.^109,110 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.¹⁰⁶

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).^111,112 After washing, labeled tracer is added and 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 obtained with this technique are generally comparable to those determined with the direct methods. They are more likely to differ in the presence of circulating inhibitors of protein binding and in sera from patients with non-thyroidal illness.

Analog (One-Step) Immunoassays. In these assays, a labeled analog of T4 or T3 directly competes with the endogenous free hormone for binding to antibodies.¹¹³ In theory, these analogs 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,^113a to transthyretin and to iodothyronine autoantibodies.¹¹⁴ This results in discrepant values to other assays in a number of conditions including non-thyroidal illness, pregnancy and in individuals with familial dysalbuminemic hyperthyroxinemia (FDH).^113a A growing number of commercial kits is available some of which have been modified to minimize these problems,^113b. Nonetheless, their accuracy remains controversial, although such comercial methods are being increasingly adopted in the routine clinical chemistry laboratory.¹¹²

Considerations in Selection of Methods for the Estimation of Free Thyroid Hormone Concentration. None of the available methods for the estimation of the free hormone concentration in serum is infallible in the evaluation of the thyroid hormone-dependent metabolic status. Each test possesses inherent advantages and disadvantages depending upon specific physiologic and pathologic circumstances. For example, methods based on the measurement of the total thyroid hormone and TBG concentrations cannot be used in patients with absent TBG due to inherited TBG deficiency. Under such circumstances, the concentration of free thyroid hormone is dependent upon the interaction of the hormone with serum proteins that normally play a negligible role (TTR and albumin). When alterations of thyroid hormone binding do not equally affect T4 and T3, discrepant results of FT4I are obtained when using labeled T4 or T3 in the in vitro uptake test. For example, euthyroid patients with the inherited albumin variant (FDH) or having endogenous antibodies with greater affinity for T4 will have high TT4 but a normal T3U test which will result in an overestimation of the calculated FT4I. In such instances, calculation of the FT4I from a T4U 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 T3U test. Similarly, use of the T4U and T3U for estimation of the free hormone concentration, is satisfactory in the presence of alterations in TBG concentration but not alterations of the affinity of TBG for the hormone.^116,117

Methods based on equilibrium dialysis are most appropriate in the estimation of the free thyroid hormone level in patients with all varieties of abnormal binding to serum proteins provided the true concentration of total hormone has been accurately determined. All methods for the estimation of the FT4 concentration may give either high or low values in patients with severe nonthyroidal illness.^{96-100,119,120} This has been attributed to the presence of inhibitors of thyroid hormone binding to serum proteins as well as to the various adsorbents used in the test procedures.^121,122 Some of these inhibitors have been postulated to leak from the tissues of the diseased patient.^123,124 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 acute changes in TBG concentration (see Chapters 5 and 16).

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 5-2). Although the free thyroid hormone concentration in serum seems to determine the amount of hormone available to body tissues, factors that govern their uptake, transport to the nucleus and functional interactions with nuclear receptors ultimately determine their biological effects.