Euthyroid hyperthyroxinemia and hypothyroxinemia

INTRODUCTION — Many conditions result in increases or decreases in serum total thyroxine (T4) and triiodothyronine (T3) concentrations, associated with normal thyroid-stimulating hormone (TSH) concentrations and no symptoms or signs of thyroid dysfunction. This constellation of laboratory values has been referred to as euthyroid hyperthyroxinemia and hypothyroxinemia, respectively. The free T4 should be normal, but many assays will report slightly abnormal values; for example, the free T4 index and many direct free T4 assays report high values in familial dysalbuminemic hyperthyroxinemia (FDH).

In the past, these conditions presented a diagnostic challenge, and many of the patients with them were inappropriately treated for thyroid disease. Today, when most clinicians measure serum TSH as a screening test for thyroid function, a normal serum TSH value is usually not followed by measurement of a total serum T4. With the use of sensitive serum TSH assays and automated "direct" free T4 assays, euthyroid hyper- or hypothyroxinemia frequently remains undetected, with no harm to the patient.

However, the detection of a normal serum TSH concentration associated with a high or low serum T4 concentration, and sometimes free T4 concentration, should immediately alert the clinician to search for one of the causes of euthyroid hyper- or hypothyroxinemia, especially if the patient has no symptoms or signs of either hyper- or hypothyroidism.

This topic will review these conditions. Individual thyroid function tests are discussed in detail separately. (See "Laboratory assessment of thyroid function".)

EUTHYROID HYPERTHYROXINEMIA DUE TO BINDING PROTEIN ABNORMALITIES — Both T4 and T3 circulate in blood bound to one of three binding proteins:

●Thyroxine-binding globulin (TBG)

●Transthyretin (TTR; thyroxine-binding prealbumin [TBPA])

●Albumin

Approximately 99.97 percent of circulating T4 and 99.7 percent of circulating T3 are bound to these proteins. T4 binding is distributed as follows: TBG, 75 percent; TTR, 15 percent; and albumin, 10 percent. In comparison, T3 is less avidly bound to TBG and TTR.

Serum total T4 or T3 assays measure both bound and free (unbound) hormone. As a result, factors that alter binding protein concentrations can have profound effects on serum total T4 (and T3) concentrations even though serum free T4 (and T3) concentrations do not change and the patient is euthyroid.

TBG excess — Thyroxine-binding globulin (TBG) excess is the most common binding protein abnormality. There are several major causes of this disorder:

●Hereditary – Hereditary TBG excess is an X-linked dominant disorder characterized by increased synthesis of TBG [1,2].

●Estrogens – Estrogens increase the glycosylation of TBG, which slows its clearance, thereby increasing serum TBG concentrations [3]. Thus, serum TBG (and total T4 and T3) concentrations are increased substantially (25 to 50 percent) in pregnant women [4], women receiving an oral contraceptive [5], postmenopausal women receiving estrogen therapy [6], and patients with estrogen-secreting tumors, and serum TBG concentrations are increased slightly (10 to 25 percent) in women receiving tamoxifen [7] and raloxifene [8].

●Hepatitis – Patients with acute or subacute hepatitis, even those with minimal elevations in serum aminotransferase concentrations, have increased serum TBG concentrations [9,10].

●Drugs – Several drugs raise serum TBG concentrations, including fluorouracil [11], perphenazine, clofibrate [12], heroin, and methadone [13].

●Acute intermittent porphyria – Acute intermittent porphyria may be associated with increased serum TBG concentrations, via an uncertain mechanism [14].

Diagnosis — The T3-resin uptake test was designed in part to detect abnormalities in serum TBG (see "Laboratory assessment of thyroid function", section on 'Serum free T4 and T3'). This test measures the number of unoccupied T4-binding sites in serum. It is performed by incubating the patient's serum with radiolabeled T3 and subsequently adding a resin (or other binder such as talc or dextran-coated charcoal) to bind any radiolabeled T3 not bound to serum proteins (figure 1). In states of TBG excess, more radiolabeled T3 binds to TBG and less to the resin, resulting in a low T3-resin uptake (figure 2).

The value in normal subjects is usually normalized to 1 (range 0.83 to 1.16). The thyroid hormone-binding ratio (THBR) or index (THBI) is the ratio of the patient's T3-resin uptake value to the mean of the values in normal subjects. The serum free T4 index, which has no units, is calculated by multiplying the serum total T4 value by the binding ratio:

The presence of TBG excess is suggested by a low THBR in a patient with an elevated serum total T4 (or T3) concentration. In a pregnant woman, for example, the serum total T4 concentration might be 14 mcg/dL (180 nmol/L), but the THBR might be 0.68. The serum free T4 index would be normal (9.5 [14 x 0.68]).

These changes are reversed in patients with TBG deficiency (see 'TBG deficiency' below). In these patients, little radiolabeled T3 binds to TBG and most binds to the resin. The net effect is that the THBR is high, leading to a normal serum free T4 index even though the serum total T4 concentration is low (figure 3).

In contrast, the T3-resin uptake and therefore the serum free T4 index are high and low, respectively, in hyperthyroidism and hypothyroidism.

The changes in T3-resin uptake in TBG excess and deficiency and in hyperthyroidism and hypothyroidism can be summarized as follows.

●Hyperthyroidism – High serum total T4, high T3-resin uptake or THBR, high serum free T4 index

●TBG excess – High serum total T4, low T3-resin uptake or THBR, normal serum free T4 index

●Hypothyroidism – Low serum total T4, low T3-resin uptake or THBR, low serum free T4 index

●TBG deficiency – Low serum total T4, high T3-resin uptake or THBR, normal serum free T4 index

At extremes of TBG excess or deficiency, the serum free T4 index may not be normal. However, the correct diagnosis can be established by demonstrating that the patient has a normal serum TSH concentration. In addition, serum TBG can be measured directly by radioimmunoassay, although this is rarely necessary.

In theory, direct free T4 assays should not be affected by changes in TBG. In practice, however, many automated free T4 assays perform poorly, especially in pregnancy and nonthyroidal illness, because they are adversely affected by the lower levels of albumin seen in pregnancy and nonthyroidal illness. Even though free T4 index measurements are normal in the second and third trimester, automated free T4 assays frequently give low values, and some kit manufacturers attempt to compensate for the poor performance of their assay by offering adjusted lower normal ranges during the second and third trimester [15]. (See "Overview of thyroid disease and pregnancy", section on 'Trimester-specific reference ranges'.)

Familial dysalbuminemic hyperthyroxinemia — Familial dysalbuminemic hyperthyroxinemia (FDH) is a genetic disorder, most often occurring in patients of Hispanic ethnicity, in whom it occurs in perhaps 0.2 percent [1,16]. It is characterized by production of mutant albumin molecules that have a low affinity but high capacity for T4, but not T3 [17]. Affected patients have high serum total T4 concentrations but are euthyroid and have normal serum TSH concentrations.

The increased binding of T4 in FDH results in a misleading T3-resin uptake because the radiolabeled T3 does not bind to the abnormal albumin but does bind normally to the resin and to other serum proteins (figure 4). The net effect is a high serum total T4 concentration and a normal THBR, so that the calculated serum free T4 index is high. Before sensitive serum TSH assays were available, these patients were commonly misdiagnosed as having hyperthyroidism and treated accordingly.

While, in theory, direct free T4 measurements should result in normal free T4 levels in FDH, in reality, most are adversely affected by changes in albumin concentration and give spuriously high values [18,19].

There is usually no need to pursue the diagnosis of FDH if the patient is clearly euthyroid and has a normal serum TSH concentration. However, the diagnosis can be established by performing a resin (or charcoal) uptake test using radiolabeled T4 rather than radiolabeled T3 [17]. Serum from patients with FDH binds more radiolabeled T4 than does serum from normal subjects, thereby documenting that serum T4 binding is increased. The diagnosis can also be established by electrophoresis of binding proteins in the presence of radiolabeled T4. Alternatively, serum T4 and free T4 index can be measured in the patient's relatives; the inheritance of FDH is autosomal dominant.

FDH can be diagnosed by assessing gene variants in the albumin (ALB) gene. R218H, R218P, R222I, and R218S variants have been described [20]. The clinical presentation of the R218P variant differed from the patients described above; serum T4 concentrations were 11- to 17-fold above the upper limit of normal (versus two- to threefold), and serum T3 concentrations were elevated up to twofold above the upper limit of normal in some family members [21,22]. This variant may also bind excess cortisol [23].

Abnormal TTR binding of T4 — This very rare disorder is due to production of an abnormal transthyretin (TTR) that, like the abnormal albumin in patients with FDH, binds T4 but not T3 [1,24]. It can be diagnosed by either electrophoresis or by using anti-TTR antibodies to immunoprecipitate the protein in the presence of radiolabeled T4 [25].

Anti-T4 immunoglobulins — The presence of anti-T4 immunoglobulins, a rare finding, can cause spuriously high serum T4 concentrations when T4 is measured by radioimmunoassay, or low values in immunometric assays [26,27]. These immunoglobulins are antibodies that bind T4, but are distinct from antithyroglobulin (and antithyroid peroxidase) antibodies. Nonetheless, virtually all patients with these antibodies have autoimmune thyroid disease. Most patients have been euthyroid because enough serum T4 was unbound to maintain normal function at the tissue level.

These human anti-T4 antibodies cause spuriously high serum total T4 values in T4 radioimmunoassays because they bind radiolabeled T4, thereby competing with binding of radiolabeled T4 to the rabbit anti-T4 antibodies used in the assays, which then interferes with the competition between labeled and endogenous T4 binding (figure 5). On the other hand, the THBR is normal because most endogenous anti-T4 antibodies do not bind T3. Some, but not all, direct free T4 assays give spuriously high results in the presence of anti-T4 antibodies [28]. Human anti-T4 antibodies cause spuriously low values in immunometric assays by preventing the endogenous T4 from interacting with the signal antibody.

Anti-T4 antibodies can be detected by adding radiolabeled T4 to the patient's serum and precipitating the immunoglobulin fraction by the addition of polyethylene glycol [25]. Precipitation of radiolabeled T4 indicates the presence of anti-T4 immunoglobulins.

Anti-T3 immunoglobulins have also been described [26].

OTHER CAUSES OF EUTHYROID HYPERTHYROXINEMIA

Reduced thyroxine deiodination — Several drugs inhibit extrathyroidal T4 deiodination to T3, including amiodarone, propranolol in high doses, and the iodinated radiographic contrast agents ipodate and iopanoic acid [29-32] (see "Amiodarone and thyroid dysfunction"). Administration of these drugs may result in hyperthyroxinemia with normal serum TSH concentrations.

Assay artifact — Many competitive binding assays for T4 or free T4 utilize biotinylated antibodies and a streptavidin-biotin separation technique. Excess ingestion of biotin competes with the biotinylated analogue and results in spuriously high T4 or free T4 (or T3 or free T3) values. Measurements should be repeated after ingestion of biotin is omitted for two days [33].

Other conditions — Several other disorders can also cause euthyroid hyperthyroxinemia:

●Acute psychosis, acting via an uncertain mechanism [34,35] (see "Thyroid function in nonthyroidal illness"). From 1 to 10 percent of patients hospitalized for acute psychosis have modestly elevated serum T4 concentrations [35]. The elevation is usually transient, but measurement of serum TSH is indicated because a few of these patients are actually hyperthyroid [35]. Still, among 12 acutely depressed patients, none had a low serum TSH concentration and seven had slightly elevated concentrations that were probably transient [36].

●High altitude and amphetamines, presumably mediated by a central nervous system mechanism [37,38].

●Symptomatic hyponatremia may be associated with small increases in serum total T4 concentrations [39].

●Generalized thyroid hormone resistance results in high serum T4 concentrations associated with normal or slightly high serum TSH concentrations. These patients may be euthyroid, or have clinical manifestations of hypothyroidism or even hyperthyroidism, because the receptor defect can vary in different organs. (See "Resistance to thyroid hormone and other defects in thyroid hormone action", section on 'Resistance to thyroid hormone beta (RTH-beta and nonTR-RTH)'.)

●Appropriate T4 replacement therapy (with normalization of serum TSH concentrations) results in serum T4 concentrations that are 1 to 2 mcg/dL (13 to 26 nmol/L) higher than in normal subjects [40].

EUTHYROID HYPERTRIIODOTHYRONINEMIA — A variant of albumin with 40-fold increased affinity for T3 but only 1.5-fold increased affinity for T4 has been described in a Thai family: familial dysalbuminemic hypertriiodothyroninemia [41].

In one study, 12.4 percent (13 of 105) newly diagnosed multiple myeloma patients, mostly of the immunoglobulin G (IgG) type, had hypertriiodothyroninemia that resolved with anti-myeloma therapy [42].

EUTHYROID HYPOTHYROXINEMIA DUE TO BINDING PROTEIN ABNORMALITIES — Changes in the opposite direction from those described above can cause euthyroid hypothyroxinemia.

TBG deficiency — Thyroxine-binding globulin (TBG) deficiency occurs in the following settings:

●Hereditary – Hereditary TBG deficiency is an X-linked recessive disorder [1,2,43]. The underlying cause is a mutation in one of the first four exons of the TBG gene [44]. In one family, however, low serum TBG concentrations were autosomally transmitted by an unknown mechanism; the patients' TBG gene was normal [45].

●Hormonal abnormalities – Androgens (high doses) reduce serum TBG concentrations and can result in a reduction in T4 requirement in hypothyroid patients [46]. Other hormonal causes of low serum TBG concentrations include glucocorticoid administration [47], Cushing syndrome, and acromegaly [48].

●Nephrotic syndrome – Urinary loss of TBG in patients with the nephrotic syndrome can result in hypothyroxinemia. Most of the patients are euthyroid, but hypothyroidism can occur in those with poor thyroid reserve or in hypothyroid patients taking T4 (levothyroxine) [49]. (See "Endocrine dysfunction in the nephrotic syndrome", section on 'Thyroid function'.)

●Drugs – L-asparaginase [50], danazol [51], and niacin [52] all lower serum TBG concentrations, presumably by decreasing TBG production.

Starvation and poor nutrition can cause reductions in both serum TBG and albumin concentrations. (See "Thyroid function in nonthyroidal illness".)

Displacement of T4 from binding proteins — Several drugs, including: salicylates [53], salsalate [54], high doses of furosemide (especially in patients with renal failure) [55], and some nonsteroidal antiinflammatory drugs (fenclofenac and mefenamic acid) inhibit the binding of T4 to TBG. As a result, serum total T4 concentrations fall, but serum free T4 concentrations are normal.

Heparin can result in spuriously low serum T4 concentrations. It does so by activating lipoprotein lipase, thereby generating free fatty acids that inhibit T4 binding to albumin [56]. The same phenomenon may occur in vivo in patients with severe nonthyroidal illness who have low serum albumin concentrations (which bind free fatty acids).

OTHER CAUSES OF EUTHYROID HYPOTHYROXINEMIA

Antiseizure medications — In euthyroid patients, both phenytoin and carbamazepine increase nondeiodinative metabolism of T4 and T3 and displace the hormones from binding proteins. As a result, serum total and free T4 concentrations decrease by approximately 40 percent; the decrease in serum T3 is smaller, but TSH concentrations remain within the normal range [57,58]. The decrease in free T4 is an artifact in most free T4 assays. Serum free T4 is usually measured in diluted serum, and dilution of the antiseizure medication diminishes its ability to displace T4 from binding proteins. As a result, some of the free T4 becomes bound during the assay, lowering the measured value [59]. Serum free T4 (and T3) concentrations are normal when measured in undiluted serum by ultrafiltration [59]. Although less well studied, oxcarbazepine has also been reported to reduce total and free thyroid hormone concentrations [60].

Thus, the thyroid status in patients treated with phenytoin or carbamazepine who have normal pituitary function should be assessed by measuring serum TSH alone. Hypothyroid patients treated with T4 may need an increase in dose after phenytoin or carbamazepine treatment is initiated [61].

Oral T3 administration — T3 is not recommended for thyroid hormone treatment in hypothyroidism (see "Treatment of primary hypothyroidism in adults"). It is, however, used by some psychiatrists as adjunctive therapy in patients with depression who respond poorly to tricyclic antidepressant drug therapy [62]. Exogenous T3 suppresses pituitary TSH production and therefore thyroidal T4 synthesis and secretion. Thus, these patients have low serum T4, normal or slightly elevated T3, and low serum TSH concentrations.

SUMMARY

●Many conditions result in euthyroid hyperthyroxinemia or hypothyroxinemia, which are characterized by increases or decreases in serum total thyroxine (T4) and triiodothyronine (T3) concentrations, associated with normal thyroid-stimulating hormone (TSH) concentrations and no symptoms or signs of thyroid dysfunction. The free T4 index may be normal or just slightly abnormal, but in some instances, it will also be increased (eg, familial dysalbuminemic hyperthyroxinemia [FDH]). Many automated direct free T4 measurements give spurious results in patients with these conditions, especially in the presence of quantitative or qualitative changes in serum albumin concentrations. Today, when most clinicians measure serum TSH as a screening test for thyroid function and most labs are using automated "direct" free T4 assays, a normal serum TSH value is usually not followed by measurement of a total serum T4. As a result, euthyroid hyper- or hypothyroxinemia may go undetected, with no harm to the patient. (See 'Introduction' above and 'TBG excess' above and 'Familial dysalbuminemic hyperthyroxinemia' above.)

●The most common cause of hyperthyroxinemia (and hypertriiodothyroninemia) is thyroxine-binding globulin (TBG) excess due to estrogens (eg, pregnancy and oral contraceptives). Other causes of TBG excess are hereditary, hepatitis, fluorouracil, perphenazine, clofibrate, methadone, heroin, and acute intermittent porphyria. FDH and other binding protein abnormalities can cause hyperthyroxinemia. (See 'Euthyroid hyperthyroxinemia due to binding protein abnormalities' above.)

●Hypothyroxinemia due to TBG deficiency is hereditary, or due to androgens, glucocorticoids, Cushing syndrome, acromegaly, nephrotic syndrome, L-asparaginase, danazol, or niacin. Hypothyroxinemia may also be due to displacement of hormone from binding proteins caused by salicylates, salsalate, high-dose furosemide, fenclofenac and mefenamic, heparin, phenytoin and carbamazepine. (See 'Euthyroid hypothyroxinemia due to binding protein abnormalities' above.)

●Hypothyroxinemia is expected with T3 (liothyronine) treatment, which reduces serum T4 levels via negative feedback on pituitary TSH production. (See 'Oral T3 administration' above.)