Unexplained high thyroid stimulating hormone: a “BIG” problem

Abstract

Macro-hormones and macro-enzymes are high molecular weight conjugates of hormones or enzymes, respectively, often with immunoglobulins. These are referred to as macromolecular complexes, and may cause artefactually elevated biochemical tests results. Macro enzymes of the most commonly measured serum enzymes have been identified and are recognised as a source of elevated measurements that may cause diagnostic confusion; macro-creatine kinase and macro-amylase are the two better known macro-enzymes in clinical practice. Literature on macro-hormones is largely restricted to macro-prolactin. We present a case of a clinically euthyroid patient, who had persistently elevated thyroid stimulating hormone (TSH) but free thyroxine within the reference limits. She underwent repeated thyroid investigations and thyroid hormone interference studies, until macro-TSH was identified as the most likely cause of unexplained elevated TSH. Following the identification and characterisation of this biochemical abnormality, she is no longer subject to repeated blood tests for assessment of thyroid function; the patient currently remains clinically euthyroid.

BACKGROUND

Macro-hormones are high molecular weight complexes, and depending on their chemical structure, may be classified as immunoglobulin bound or non-immunoglobulin bound. There is evidence that due to their large size, macro-hormones are less efficiently cleared from the circulation by the kidneys and consequently tend to accumulate in the serum, sometimes reaching very high concentrations. They are rarely associated with disease states, and, unless identified, may be a source of misleading high results.¹ This has the potential to lead to unnecessary repeated clinical investigations or management. This case demonstrates the importance of identifying a macro-thyroid stimulating hormone (TSH) in a patient with an unexplained elevated TSH concentration, which prevented further unnecessary investigations.

CASE PRESENTATION

A 46-year-old woman, not on any treatment, presented with diffuse pain in both upper arms without any loss of function. There was no history of weakness, weight gain, loss of appetite, undue cold intolerance or menstrual irregularities. There was no family history of thyroid dysfunction. Systemic examination was unremarkable; the thyroid gland was not enlarged or tender. She underwent “routine” blood tests including TSH and free thyroxine (fT4), which revealed high TSH with fT4 within the reference limits.

INVESTIGATIONS

Over the next 3 months she had repeat blood tests to assess thyroid function (table 1A). TSH was consistently elevated, while the fT4 was within the reference limits; thyroid peroxidise antibodies (TPO) were not raised. Since she was clinically euthyroid and not on thyroxine replacement therapy, she underwent further investigations to determine the possible cause of the unexpected high TSH. To exclude a method specific interference, TSH and fT4 were measured on a different analytical platform by a different immunoassay. The results confirmed the elevated TSH and “normal” fT4 (table 1B). The sample was tested in doubling dilutions for potential analytical interference. It did not dilute linearly (non-parallelism), which was suggestive of the presence of interfering antibodies (table 2).

Table 1.

Thyroid function tests using different methods on different analytical platforms

Analysis platform	TSH (mU/l)(0.4–4)	fT4 (pmol/l)(9.9–22)
(A) Roche (E170)*	24.5	16.8
	38.1	15.9
	37.2	15.5
(B) Bayer (Centaur)†	TSH (mU/l)(0.3–3.5)	fT4 (pmol/l)(11.8–24.6)
	18	14.7

*Roche Diagnostics Limited, Burgess Hill, West Sussex, UK.

†Bayer plc, Strawberry Hill, Newbury, Berkshire, UK.

Table 2.

Non-linearity of thyroid stimulating hormone (TSH) on dilution

Sample	TSH (mU/l)
Undiluted	38.1
1:12 dilution	26
1:144 dilution	14.8

The presence of human anti-animal antibodies in the test sample may cause interference in immunoassays and lead to a false positive result.² The patient’s sample was treated with reagents containing anti-mouse, anti-sheep and anti-rabbit blocking antibodies. TSH was measured before and after treatment with the blocking antibodies—there was no significant change in TSH, which did not support the presence of interference by human anti-animal antibodies.

At this point an aliquot of the serum sample was sent to the research and development department at Roche laboratories (Roche Diagnostics Limited, Burgess Hill, West Sussex, UK), where further analysis was undertaken to try to determine the cause of the elevated TSH (table 3). TSH and fT4 were confirmed on two analytical platforms. Heterophile antibodies are non-specific and generally weak antibodies against poorly defined antigens. To test for the presence of heterophile antibodies, the sample was tested before and after treatment with a heterophile antibody blocking reagent. The Roche assay uses antibody labelled with a ruthenium complex, antibodies against which have been demonstrated to cause analytical interference.³ The sample was treated with ruthenium interference blocking protein before measuring TSH. There was no demonstrable evidence of either heterophile or anti-ruthenium antibodies in the patient’s sample. Finally, size exclusion chromatography was performed. This technique separates analytes according to their molecular size. Unexpectedly, TSH was found to elute with the immunoglobulin G (IgG) fraction, thus confirming the presence of a macromolecular complex of TSH-IgG (macro-TSH).

Table 3.

Interference studies carried out at Roche laboratories

Analysis platform	TSH (mU/l)(0.3–3.5)	fT4 (pmol/l)(11.8–24.6)	fT3 (pmol/l)(3.0–7.8)
(A) Bayer (Centaur)	14.5	18.7	4.7
(B) Roche (Elecsys)	TSH(0.27–4.2)	fT4(12–22)
(i) Untreated sample	38.72	18.1
(ii) Treated with Fab2	35.84
(iii) Treated with sRu	39.15

fT3, free tri-iodothyronine; fT4, free thyroxine; TSH, thyroid stimulating hormone.

Fab2 is a heterophile antibody blocking protein; sRu is a ruthenium interference blocking protein.

OUTCOME AND FOLLOW-UP

This patient remains clinically euthyroid; biochemically, TSH is persistently elevated and the fT4 is within the reference limits. The presence of macro-TSH, in the absence of clinical features suggestive of a thyroid disorder, has prevented further unnecessary investigations (especially for a TSH secreting pituitary adenoma) or a trial of thyroxine replacement. In patients like this thyroid function tests should always be interpreted in the light of clinical symptoms and previous test results and not population based reference intervals.

DISCUSSION

Unexplained elevated TSH without a coexisting low fT4, in the absence of clinical features suggestive of hypothyroidism or thyroxine replacement, is often attributed to non-thyroid illness (NTI), some sort of interference in the analysis of thyroid hormones, or subclinical hypothyroidism. By definition, subclinical hypothyroidism encompasses high TSH, normal fT4 and absence of clinical features of hypothyroidism. However, a past history of thyroid disease and thyroxine treatment and other causes of a raised TSH, such as a TSH producing pituitary adenoma, need to be excluded before a diagnosis of subclinical hypothyroidism may be made.⁴ The accepted upper limit of TSH is usually 10 mU/L, which is consistent with subclinical hypothyroidism.⁵ A higher value (as present in our patient) is almost always considered pathological and necessitates further investigations. The molecular weight of TSH is approximately 30 kDa and it is easily filtered by the kidney. However, on combining with an immunoglobulin (Ig) G molecule, the molecular weight of this large complex increases to approximately 200 kDa, which precludes its filtration by the kidney and leads to accumulation of macro-TSH in the serum. This TSH–IgG complex is biologically inactive, but remains immunoreactive (hence measurable). It undergoes degradation, in an unpredictable fashion, resulting in variable TSH concentrations at different times. This may also explain why TSH analysis produces significantly different results on different analytical platforms.

In general, macro-hormones and macro-enzymes are well recognised, at least by laboratory staff. However, macro-TSH is an uncommon form of a macromolecular complex, and the paucity of literature on it suggests that it is not widely recognised or even considered as a possible cause of unexplained high TSH. Three cases of macro-TSH were described in the literature more than two decades ago.^6–9 A fourth case was described in 2006,¹⁰ where macro-TSH in a neonate resulted in a false positive screening test result for congenital hypothyroidism, which prompted investigation of both the baby and mother for macro-TSH. In this case, the macro-TSH in the baby was considered to be of maternal origin.

The prevalence of macro-TSH remains speculative and it is likely that most cases of macro-TSH remain unidentified. It is also possible that a proportion of patients diagnosed as having subclinical hypothyroidism, who never become hypothyroid, may harbour a macro-TSH.¹¹

LEARNING POINTS

• Elevated TSH with normal fT4 is a common finding and may be due to many different causes.

• Macro-TSH may lead to artefactual TSH elevation, which is not a widely recognised phenomenon, and may result in unnecessary investigations or trial of thyroxine treatment.

• The prevalence of macro-TSH is unknown; it is quite likely that a number of patients labelled as having subclinical hypothyroidism, may harbour macro-TSH.

• Once a macro-hormone (macro-enzyme) complex has been identified in a patient, future results should be interpreted in the light of clinical features and previous test results, rather than population based reference intervals.

• Laboratory test results should always be interpreted in light of clinical findings, and a close liaison maintained between clinicians and laboratory personnel.