Diagnostic Dilemma: A Case of Undetectable TSH
Abstract Introduction: Under steady-state conditions, measurement of TSH is accepted as the best assessment of thyroid function. The widely used TSH chemiluminometric assays have very low limits of detection and can help distinguish between the various causes of subnormal TSH. However, when evaluating a patient with abnormal thyroid tests but without thyroid symptoms, an appraisal of the test should be considered. Clinical Case: A 63-year-old South Asian man was referred to endocrinology for evaluation of a non-detectable TSH (<0.01 µIU/mL) that was reproduced on repeat testing, both using Siemens ADVIA Centaur TSH3-UL immunoassay. The patient was clinically euthyroid and denied taking biotin supplements. Testing of thyroid hormone showed normal values for free T3, total T3, free T4, and total T4. Additional labs included normal studies for free thyroxine by equilibrium dialysis, thyroid stimulating immunoglobulin, and heterophilic antibodies. Thyroid uptake and scan showed uniform uptake of 5.1% and 15.1% at 2-hours and 24-hours, respectively, with no dominant nodules. Hypothalamic-pituitary hormonal testing and MRI pituitary were both normal as well. When TSH testing was repeated on a separate platform, Roche’s eCLIA immunoassay, detectable values were obtained (TSH 6.48 µIU/mL). Conclusions: Testing of serum TSH by commercially available immunoassays is based on the sandwich method in which one antibody binds to the β-subunit of TSH and the other to the α-β interface. Most assays use monoclonal antibody pairs to achieve high selectivity. Immunoassay tests are prone to interferences, particularly by way of altering the measurable concentration of the analyte or by altering antibody binding (1). In this case, the presence of detectable TSH depended on the platform by which it was measured. This finding suggests a TSH-β variant with impaired immunoreactivity but functionally normal bioactivity. Such a mutation has been previously reported to occur five times more frequently among South Asian individuals than the general population (2). Genetic testing was offered to the patient to confirm this suspicion but was declined. It is incumbent on the clinician to reconcile a test result that is discordant with the clinical presentation. Having a fundamental understanding of the principles of the testing platform can assist in identifying potential sources of error. Failing to recognize a possible interference can lead to unnecessary healthcare expenditures, misdiagnosis and inappropriate management, potentially at a cost to the patient’s wellbeing. When faced with an undetectable TSH with otherwise normal thyroid hormones and unremarkable clinical picture, it is best to repeat the TSH test using a different available platform. References: (1)Favresse J et al. Endocr Rev. 2018;39(5):830-850(2)Pappa T et al. Thyroid. 2015 Aug;25(8):869-76
