Thyroid Testing

     Few laboratory tests are as difficult to interpret as thyroid function tests, in part because tests for thyroid include measurement of a protein synthesized in another organ, the liver. This protein is thyroxine binding globulin (TBG) and is the major carrier protein for thyroxine. Pre-albumin and albumin, the other carrier proteins, are of minor significance. Diseases of the thyroid have little or no effect on TBG, yet it is a necessary component of thyroid testing. Apart from a minute fraction, 0.03% which is free, 99.97% of thyroxine circulates in its bound form; hence any change in TBG will result in a change in circulating total thyroxine. A number of non-thyroidal conditions have profound effects on circulating levels of TBG and thus total thyroxine. The free thyroxine, on the other hand, reflects a patient's thyroid status and is generally independent of TBG and any effect of TBG on total thyroxine levels. Therefore, the three components of thyroid function are total thyroxine, an assessment of thyroxine binding proteins (mainly TBG), and free T₄. Free T₄ can be calculated as an estimated free thyroxine (EFT) or expressed indirectly as the free thyroxine index (FTI) from the total thyroxine and thyroxine binding proteins.

1. Total Thyroxine (T₄) - All current measurements of T₄ are immunoassays. The method used at Yale is a fluorescence polarization immunoassay. Total thyroxine is increased in hyperthyroidism or in hyperthyroxinemia due to a variety of causes. Hyperthyroxinemia is defined as increased total thyroxine not necessarily due to hyperthyroidism (Grave's disease). Hyperthyroxinemia may be seen in the initial phases of thyroiditis, with release of T₄ from a damaged gland, or in cases of the stress of illness, but to a lesser extent. Increased total T₄ due to increases in TBG is seen in pregnancy, after estrogen therapy, in hepatitis, and in a variety of non-thyroidal illnesses. In dysalbuminemic hyperthyroxinemia, total thyroxine is increased secondary to the presence of a thyroxine binding, albumin-like protein. Decreased total thyroxine occurs in hypothyroidism, primary or secondary, and in non-thyroidal stress situations such as illness. Decreased total T₄ due to a decrease in TBG may be seen in congenital decreases in TBG, acquired decreases in TBG such as severe liver disease, and in the nephrotic syndrome.

   As an illustration of this complicated physiology, consider a pregnant individual or an individual on estrogen who has a "normal" total thyroxine of 6 µg/dL (reference range 5-10.6 µg/dL). Could this patient be hypothyroid? Should the total thyroxine be higher because of the increased TBG due to estrogen effect? Because of the effect of TBG on circulating total thyroxine, total thyroxine should never be used alone to evaluate thyroid disease. The two major interferences in the total T₄ assay are auto-antibodies against thyroxine and the peculiar T₄ binding protein in dysalbuminemic hyperthyroxinemia. In the former, the result may be falsely high, falsely low, or even spuriously normal depending on the mechanics of the particular assay used for total thyroxine. For the latter, total T₄ will be elevated but this is the true state of affairs and reflects increased protein binding capacity for thyroxine. Patients with dysalbuminemic hyperthyroxinemia are euthyroid.

2. Tests for the assessment of Thyroxine binding proteins, primarily TBG - Most tests are functional assays that assess the ability of the patient's serum to bind exogenously introduced thyroxine. The concentration of the TBG protein, as well as the degree of endogenous thyroxine saturation on TBG, will be reflected in the value obtained by these "T uptake" tests. At Yale, we use the T uptake value to calculate the thyroxine binding capacity (TBC). The only purpose of obtaining either the T uptake or the TBC is to enable calculation of the FTI or the EFT. The T uptake test results are usually inversely related to the degree of unsaturation of TBG. For example, if the patient's TBG is more saturated with endogenous thyroxine, with fewer unsaturated sites, as in hyperthyroidism, the T uptake result will usually be elevated. The TBC is directly related to the degree of unsaturation; in hyperthyroidism the TBC would usually be decreased. In hypothyroidism, the converse would hold; the T uptake will usually be decreased and TBC usually increased. Since these tests reflect both the degree of thyroxine saturation on TBG and the amount of TBG protein, the value is a composite and therefore the results do not always go in the expected direction. As is the case with total thyroxine, tests for binding proteins, T uptake and TBC, should not be used alone to evaluate thyroid disease. Their purpose is to enable the estimation of a free thyroxine index in the case of T uptake or an estimated free thyroxine in the case of TBC.

3. Calculation of the Estimated Free Thyroxine (EFT) or Free Thyroxine Index (FTI). These reflect the free thyroxine concentration in the blood. This is the metabolically active fraction which best reflects the status of the thyroid gland and is generally not appreciably influenced by changes in binding proteins. This fraction is termed the estimated free thyroxine (EFT) at Yale and the free thyroxine index (FTI) at the West Haven Veterans Hospital. The EFT is given in ng/dL of free thyroxine whereas the FTI is a dimensionless number. Both calculations are based on the reversible equilibrium relationship of the following mass action equation:
   

   where:
   = free thyroxine
   = unsaturated binding sites on thyroxine binding proteins, mainly TBG
   = total Thyroxine bound to Thyroxine binding proteins, mainly TBG
   = equilibrium constant for the mass action equation

   The question usually asked is whether the level of total thyroxine is appropriate to the functional level of thyroxine binding proteins. A result above the reference ranges for EFT or FTI indicates hyperthyroxinemia and may indicate hyperthyroidism; a result below the reference range is hypothyroxinemia and may indicate hypothyroidism. The formal diagnosis of hyperthyroidism or hypothyroidism depends on further testing. The farther the values deviate from the reference ranges, the more likely is true thyroid disease, but further workup is usually necessary. What about direct measurements of free thyroxine? Commercial systems have recently become available for the measurement of free thyroxine directly by immunoassay and are designated as analog one or two step assays. The direct measurement of free thyroxine is inherently difficult because the aim is to measure 0.03% of free thyroxine against a backdrop of 99.97% bound to protein in serum. The mechanics of the assays involve capturing the free T₄ by antibody without causing a shift in the reversible equilibrium of the mass action equation defining free T₄. This is a formidable task. Furthermore various interferences causing spurious results in direct free T₄ measurements have been reported including rheumatoid factor, heterophile antibodies, anti-T₄ antibodies, and very low or very high TBG levels (1,2). Drugs that compete with T₄ for TBG binding sites could be potential sources of error. Moreover, some direct assays do not give an accurate estimate of free thyroxine in non-thyroidal illness. Direct measurements of free T₄ may be useful in many patients but the exceptions are significant and worrisome. The American Thyroid Association has gone on record as having serious reservations about direct methods for measuring free T₄ (3). Furthermore, they are not necessary for the accurate diagnosis of thyroid disease since the EFT and FTI perform very well clinically and have done so for several years. However, interferences do occur here also. For example, in the usual direct free T₄ assays (with the exception of the free T₄ assay by equilibrium dialysis) and in the calculated EFT or FTI, the results in dysalbuminemic hyperthyroxinemia may be spuriously high, despite a euthyroid state. Oftimes having a result for total thyroxine and for binding proteins, rather than just a value for direct free T₄, may provide a clue about potential interference in the tests.

Thyroid Stimulating Hormone (TSH)
     Over the past five years or so, high sensitivity and very high sensitivity assays for TSH have been introduced (second and third generation assays, respectively). Whereas in the past, TSH was used to verify the diagnosis of hypothyroidism, it could not be used reliably for hyperthyroidism because the assays were not sensitive enough in the low range to detect truly suppressed levels. Current assays make this possible and TSH is now a valuable adjunct to thyroxine studies in the diagnosis of hyperthyroidism.

     In cases where the EFT or FTI are low, the TSH is very helpful. In cases of primary hypothyroidism, the TSH is elevated whereas in the hypothyroxinemia of the euthyroid sick syndrome, TSH is generally normal. In hyperthyroidism (Grave's disease), the TSH is suppressed to <0.04 µIU/mL in the second generation assay and to <0.002 µIU/mL in the third generation (at Yale). It should be noted that in the euthyroid sick syndrome a TSH will be occasionally suppressed in the second generation to <0.04 µIU/mL but will be measurable by the third generation. Since cortisol and steroids in general suppress TSH, this may be secondary to increased secretion of cortisol during stress. Occasionally, a mild increase in TSH occurs in the euthyroid sick syndrome, as well as mild hyperthyroxinemia and mild increases in free thyroxine, estimated or directly measured. Alterations in thyroid function tests that occur during non-thyroidal illness usually resolve when the patient recovers. TSH results should also be interpreted with caution in patients who are receiving dopamine or steroids since these suppress TSH secretion. Patients who are treated for hyperthyroidism may have suppressed TSH for weeks or even months after thyroxine and free thyroxine levels return to normal or even fall into the hypothyroid range. Patients with dysalbuminemic hyperthyroxinemia will have normal TSH results.

Triiodothyronine (T₃)
     Although T₃ is one of the hormones produced by the thyroid, approximately 80% of circulating T₃ is the result of peripheral deiodination of T₄. Many non-thyroidal situations affect T₃, generally decreasing levels. These include non-thyroidal illnesses, various drugs such as glucocorticoids and propranolol, and caloric deprivation. Except in very special circumstances, T₃ should not be used to assess thyroid function. It may be indicated if there is suspicion of "T₃ thyrotoxcosis" where the free thyroxine is generally high normal with concomitant suppression of TSH. Also, after radioactive ablation therapy for Grave's disease, the T₃ may be elevated with a normal T₄. T₃ may also be helpful in assessing the patient for dysalbuminemic hyperthyroxinemia. Although total T₄ is elevated total T₃ is normal because the albumin-like protein in dysalbuminemic hyperthyroxinemia does not bind T₃.

Thyroglobulin
     Thyroglobulin, the matrix protein in the thyroid gland, is elevated in a variety of thyroidal conditions but is used primarily in the management of patients with thyroid cancer; it is used for the detection of disease recurrence. It should be noted that a significant number of patients develop anti-thyroglobulin antibodies which interfere with the assay for thyroglobulin. Thus, it is mandatory that sera for thyroglobulin be simultaneously tested by high sensitivity tests for anti-thyroglobulin antibodies. The presence of antibodies prevents the result for thyroglobulin from being reliably used. Commercial laboratories do, however, report thyroglobulin values in the presence of antibodies but provide a disclaimer on the report.

Anti-Thyroid Peroxidase Antibodies
     These antibodies are elevated in a variety of thyroidal conditions, including Hashimoto's thyroiditis and Grave's disease.

Thyroid Stimulating Immunoglobulin (TSI)
     The basic defect in Grave's disease is an immunoglobulin which binds to the TSH receptor on the thyroid gland mimicing the action of TSH, and thereby causing hyperthyroidism. The clinical indications for TSI are few since Grave's disease can be readily diagnosed clinically and with routine tests of thyroid function. Increasing titers of TSI are thought by some to be predictive of recurrence of hyperthyroidism after therapy, but the test is not routinely advised for this purpose. TSI does cross the placenta and may cause transient hyper- thyroxinemia in the fetus and therefore the test may be indicated in pregnant women with a history of Grave's disease.

Suggested Algorithm for the Workup of Thyroid Disease
     Thyroxine indices, which includes total T₄, TBC, and EFT at Yale or Total Thyroxine, T uptake, and FTI at other institutions should be obtained first. If the EFT or FTI are well within the reference range, then the workup can stop here in most cases unless the index of suspicion is high for thyroid disease. If the EFT or FTI are borderline high or borderline low, then a serum second generation TSH assay may be in order. If the EFT or FTI are frankly elevated, then either a second or third generation TSH assay should be obtained, preferably a third generation with the greatest sensitivity at the low end. If the EFT or FTI are low, then a second generation TSH assay is in order to establish hypothyroidism.

     With the advent of the highest sensitivity TSH assay (third generation) some people feel that this could serve as the initial thyroid screening test. The reasoning is that a normal TSH rules out thyroid disease, a suppressed TSH (<0.002 µIU/mL at Yale) indicates possible hyperthyroidism, and an elevated TSH indicates possible hypothyroidism. In the latter two cases, confirmatory thyroxine studies would then follow. This is not an unreasonable scenario but the clinical question is still directed at the level of thyroid hormone rather than the level of TSH. The symptoms and signs of thyroid disease are due to thyroxine, not TSH. Although secondary hypothyroidism (pituitary deficiency of TSH) is uncommon, it does occur and can be missed by a TSH alone since TSH may be "normal" in this situation. As noted above, exogenous steroids and dopamine or L-dopa may suppress an elevated TSH to normal levels. There have also been reported cases of positive interferences in the TSH assay due to anti-murine antibodies in patients giving a higher than expected result in known cases of Grave's disease (TSH in the "normal" range rather than being suppressed) (4).

     In the situation where thyroxine studies are normal or high normal and TSH is suppressed to <0.002 µIU/mL, then a serum T₃ may be in order to rule out T₃ thyrotoxicosis. This is uncommon in the U.S. but does occur more frequently in areas of the world with iodine deficiency. Similarly in post-irradiation therapy for hyperthyroidism, a T₃ may be useful if thyroxine studies are normal or low and TSH is suppressed, since T₃ may be elevated. However, one must keep in mind that even with normal or low EFT or FTI and normal or low T₃ the TSH may remain suppressed post-therapy (either by drug or radioactive ablation) for significant periods of time. For further information regarding thyroid testing, please contact Dr. Richard Donabedian at 688-2445.

Richard Donabedian, M.D.

References

1. Norden AG, Jackson RA, Norden LE, Griffin AJ, Barnes MA, and Little JA. Misleading results from immunoassays of serum free thyroxine in the presence of rheumatoid factor Clinical Chemistry, 43:957-62, 1997.

2. Textbook of Clinical Chemistry, 3rd Ed., Tietz NW, Burtis CA (Ed), Ashwood ER (Ed), W.B. Saunders Co., p. 1521, 1998.

3. Hay, I.D., Bayer, M.F., Kaplan, M.K. et al. American Thyroid Association assessment of current free thyroid hormone and thyrotropin measurement and guidelines for future assays. Clin. Chem. 37:2002-2008, 1991.

4. Frost SJ, Hine KR, Firth GB, Wheatley T. Falsely lowered FT₄ and raised TSH concentrations in a patient with hyperthyroidism and human anti-mouse monoclonal antibodies. Annals of Clinical Biochemistry, 35:317-320, 1998.