Discordant measurements of serum triiodothyronine (T4),1 thyroxine (T3), and thyroid-stimulating hormone (TSH) should always raise suspicions for unusual conditions. The clinical condition of the patient is especially relevant and should help guide further laboratory investigation. In the vast majority of instances, when a patient has authentic clinical hyperthyroidism with an increased T4 or T3, serum TSH should be low or undetectable (i.e., <0.01 mIU/L). Van Der Watt et al. nicely describe the possible causes of an increased serum free T4 (FT4) with a normal FT3 and TSH. I would like to emphasize several specific circumstances. A TSH-secreting pituitary tumor can mediate hyperthyroidism although serum TSH is low or normal (1). TSH bioactivity depends on proper glycosylation of the TSH molecule, and this may be altered in patients with a TSH-secreting pituitary tumor or a nonfunctional pituitary tumor (1). In these circumstances, measured TSH may be discordant from TSH bioactivity. TSH-secreting pituitary tumors may also disproportionately secrete α subunit of TSH compared with the entire TSH molecule (1).
If discordant FT4, FT3, or TSH values are obtained, it is appropriate to repeat the measurements in several different assays. Measurement of total T4 and T3 may also be helpful. The degree of laboratory abnormality suggests assay interference. In the case described by van der Watt et al., FT4 was markedly increased in the context of normal FT3 and TSH. The coexistence of another autoimmune disorder, subacute cutaneous lupus erythematosus (as well as Hashimoto’s thyroiditis) also suggests there may be specific T3 or T4 antibodies formed. Finally, the performance of additional biologic testing such as a radioactive iodine uptake may be helpful.
In the case report of Kellogg et al., a young woman with attention deficit hyperactivity disorder (ADHD) had increased serum total T4 and total T3, with low normal TSH. FT4 determined by direct dialysis and RIA were normal. They investigated the patient’s serum sample and determined there was antibody interference, although the precise nature of the antibody could not be identified. Several issues should be further emphasized.
Discordant thyroid function tests can occur due to antibody interference as described. The occurrence of discordant thyroid function tests in a young woman with a history of ADHD does raise the possibility of thyroid hormone resistance (2).
Thyroid hormone resistance (which is relevant to the patients in both the van der Watt et al. and Kellogg et al. case reports) usually occurs as a result of a molecular defect in the T3 receptor that decreases the ability of the receptor to bind T3, resulting in impaired T3 action in the periphery as well as the pituitary level (3). These patients may be clinically euthyroid despite having increased T4 and/or T3 and detectable TSH (3). Family members may have similar findings, helping to confirm the diagnosis; clinically it is difficult to sequence the T3 receptor, as the test is not readily available commercially.
Although the vast majority of cases of thyroid hormone resistance are due to nuclear T3 perturbations, recently Friesema et al. (4) have described patients with altered membrane thyroid hormone transport proteins. They identified inactivating mutations of the monocarboxylate transporter 8 and associated severe clinical abnormalities of psychomotor retardation and discordant thyroid hormone concentrations.
Kellogg et al. also discuss the ability of heterophilic antibodies to interfere in thyroid hormone and TSH assays. These antibodies can interfere with total as well as free thyroid hormone measurements and are found most commonly in individuals with autoimmune disorders or who have worked with animals.
It has been thought that the quickest method to help determine whether there are interfering substances in a thyroid hormone assay is to dilute the sample and determine if the measured concentrations in the diluted sample are linear and parallel with those for a standard or another serum sample. Kellogg et al. point out that this technique may be insensitive and may not be applicable to all assays. More efficient and direct laboratory measures (e.g., addition of blocking immunoglobulins and alternate assay measurements) should be used when discordant results are obtained.
The case reports of van der Watt et al. and Kellogg et al. nicely emphasize the clinical importance of recognizing and evaluating circumstances that can result in discordant values of T4, T3, and TSH.
Grant/Funding Support: Supported by a thyroid cancer study, Pfizer Inc.
Financial Disclosures: None declared.
↵1 Nonstandard abbreviations: T4, triiodothyronine; T3, thyroxine; TSH, thyroid-stimulating hormone.
- © 2008 The American Association for Clinical Chemistry