Case report

A small nodule in the lower left thyroid lobe was detected in an otherwise healthy 64-year-old woman at routine medical examination. Her only medical history of note was an episode of hepatitis 30 years previously. Thyroid function tests (thyroid-stimulating hormone 0.7 mU/l, free thyroxine 15 pmol/l) and a thyroid radioisotope scan were normal. FNAC showed only benign epithelial thyroid cells and scant giant cells although there was some suspicion that amyloid may be present.

Plasma calcitonin concentration (210 ng/l, normal <80 ng/l), measured by a radioimmunoassay (RIA, polyclonal rabbit anti-human calcitonin antibody with second anti-rabbit IgG antibody precipitation), was lower than generally seen in MTC but over the next 6 months rose progressively. Pentagastrin stimulation produced a peak response (790 ng/l) 76% above the basal value (450 ng/l). The patient underwent a left hemithyroidectomy from which her recovery was uneventful. Histology showed no evidence of MTC: there was microfollicular proliferation but no significant C-cell hyperplasia and staining for chromogranin was negative.

Postoperatively, her plasma calcitonin remained persistently increased (390–750 ng/l) but when measured by a two-site immunoradiometric assay (IRMA, Medgenix, Belgium), was within the reference range for that assay on several occasions (5.3–11.2 ng/l, normal <11.8).

Plasma from this patient was lyophilized and distributed to laboratories participating in the UK National External Quality Assessment Scheme (UKNEQAS) for calcitonin, five of whom were using RIA methods and six immunometric assays (IMA). Although only one laboratory produced a result which exceeded its stated reference range, large differences were seen between the results produced by RIA methods (mean 130, median 64, range 4.1–440 ng/l) and IMA methods (all results <11 ng/l). Overall there was a more than 200-fold difference in the concentrations of calcitonin reported.

Discussion

The finding of a slightly elevated plasma calcitonin concentration in a patient with a thyroid nodule but negative FNAC and no family history of MTC presents a not uncommon diagnostic conundrum.5 Standard clinical practice in such circumstances is to perform a pentagastrin stimulation test. However, the normal response to pentagastrin stimulation is not well established5; sex-related differences in response (peak male response 2.9 times higher than female) are not well appreciated6; additionally, Marsh et al 6 have shown that, in individuals from informative families who are negative for the familial MTC-causative RET proto-oncogene mutation, pentagastrin stimulation commonly produces false positive results. Often, as seen here, pentagastrin stimulation is not elucidatory in such cases.5

Elevated serum calcitonin has been reported in renal failure but renal function was normal in the present case (urea 2.7 mmol/l, creatinine 70 μmol/l). Non-thyroidal tumours may secrete calcitonin, particularly tumours of the breast, lung and neuroendocrine system. However, there was no evidence to suggest an ectopic source of calcitonin: bilateral mammography and a computed tomography scan of the chest and abdomen did not reveal any abnormality. Urinary excretion of adrenaline (0.06 μmol/24 h, normal <0.1), noradrenaline (0.41 μmol/24 h, normal <0.57), dopamine (0.92 μmol/24 h, normal <2.5), homovanillic acid (18 μmol/24 h, normal <44), 4-hydroxy-3-methoxy mandelic acid (19 μmol/24 h, normal <35), 5-hydroxyindole acetic acid (16 μmol/24 h, normal <42) and serum chromogranin A (11 U/l, normal <50) were all within laboratory reference ranges.

In the present case, the situation was complicated by the presence of a form of calcitonin which was detected in a standard RIA but not by an assay using antibodies directed against two sites on the molecule (so-called ‘sandwich’ assays). Immunoreactive calcitonin in the plasma of both normal subjects and patients with MTC has been shown to be heterogenous, probably reflecting the presence of the prohormone or intermediate forms of the prohormone and fragments of the mature molecule in the circulation.7 These peptides may be recognised by the antibodies employed in some assays: interference from calcitonin-like substances in healthy individuals has been documented as a relatively rare occurrence in calcitonin RIAs.8 Poor inter-laboratory agreement is a common feature of the UKNEQAS calcitonin scheme,9 reflecting calibration differences, non-specific background interference in RIA methods and differing specificities of the antibodies: certainly two-site IRMAs have in some cases been shown to be more selective for calcitonin than RIA,10 but sandwich assays are used by less than half the laboratories in the UKNEQAS scheme and are not widely available in the USA.4

MTC is a relatively uncommon condition which may present to either endocrinologists or surgeons. Although measurement of calcitonin tends to be performed in specialist units in the UK, treatment and management of the disease is often in general hospitals, where clinicians and laboratory scientists may be unaware of the lack of consensus in calcitonin measurement, the problems created by cross-reactivity with calcitonin-like peptides and the difficulties associated with interpretation of the pentagastrin test. In the present case, these analytical problems resulted in a patient being subjected to the inconvenience of a series of diagnostic procedures and unnecessary surgical risk.

The diagnostic problem represented by this case will become more common if recommendations are introduced to screen all patients with a thyroid nodule for MTC using plasma calcitonin measurements, based on the results of studies using two-site assays. Although there now appears to be good evidence that this approach is warranted in patients with thyroid nodules, this recommendation should only be applied when specific calcitonin assays are used.