Diagnostic Testing for Graves' or Non-Graves' Hyperthyroidism

A Comparison of Two Thyrotropin Receptor Antibody Immunoassays With Thyroid Scintigraphy and Ultrasonography

Lorenzo Scappaticcio; Pierpaolo Trimboli; Franco Keller; Mauro Imperiali; Arnoldo Piccardo; Luca Giovanella


Clin Endocrinol. 2020;92(2):169-178. 

In This Article

Materials and Methods

Study Design and Patient Recruitment

We conducted a retrospective analysis of consecutive patients who received a diagnosis of thyrotoxicosis at Competence Thyroid Centre ('Ente Ospedaliero Cantonale', Bellinzona and Lugano – Switzerland) from 2014 to 2018. From medical records included in hospital data sets, we gathered information of untreated patients with thyrotoxicosis at their first examination. Laboratory results, scintigraphy and US images of medical details of an individual patient were collected in different softwares, so that they could be only accessible to the doctor separately. Patients were being referred to our multidisciplinary team as they developed clinical manifestations of thyrotoxicosis (mainly cardiovascular and neurological symptoms) or, more rarely, to investigate incidental subnormal thyrotropin (TSH). It needs to be specified that in our Centre patients with thyrotoxicosis routinely underwent TRAb measurement, thyroid ultrasonography and scintigraphy in the same day or within few days (few weeks at the most) from the first examination. Moreover, from December 2014 to the first half of 2018, for each thyrotoxic patient, our institutional laboratory completed the thyroid evaluation with TSI measurement. Newly diagnosed and untreated patients with biochemically confirmed diagnosis of thyrotoxicosis/hyperthyroidism (ie TSH <0.1 mIU/L with normal (subclinical) or elevated thyroid fT4 and/or fT3 alone [overt]) fulfilling the following criterion were enrolled in the current study: (a) their diagnostic exams at initial workup included thyroid functional tests, both TRAb and TSI measurements, thyroid scintigraphy and ultrasonography. Patients were excluded if (a) they did not undergo all the above diagnostic exams; (b) they had previous history of thyroid disease or using L-T4 or T3 formulations; (c) they were pregnant; and (d) they had severe hepatic or renal dysfunction.

Thyroid Functional Tests and Antibody Assays

Thyroid function (including TSH, free-triiodothyronine [fT3], free-thyroxine [fT4], fT3/fT4 ratio) was assessed with immunoassay method in automated platform (Elecsys® e601 Roche Diagnostics) present at our hospital laboratory. TSH assay had analytical sensitivity of 0.006 mIU/L, and reference range was between 0.40 and 4.00 mIU/L. Normal reference ranges for fT3 and fT4 were 3.1–6.8 pmol/L and 12–22 pmol/L, respectively. Baseline diagnostic thyroid results were extracted for the present analysis.

In our institution, a commercial 2nd generation TSH receptor autoantibody immunoassay based on time-resolved amplified cryptate emission (TRACE) technology (Thermo Scientific BRAHMS Kryptor® compact PLUS, hereafter referred as Kryptor®) was used for TRAb measurements. The laboratory lower cut-off for a positive sample was ≥1.5 IU/L. The functional sensitivity was 0.3 IU/L. TRAb values exceeding 100 IU/L were reported by the laboratory as >100 IU/L. The method conformed to the 1st IS, NIBSC Code 90/672.

Immulite® TSI is a fully automated chemiluminescent immunoassay claimed to directly detect TSI. It employed a pair of recombinant human TSH receptor (TSHR) constructs in a bridging format: the capture and the signal receptor.[12] The capture construct was a TSHR chimera with the N-terminus of human TSHR binding the stimulating TRAb (S-TRAbs) and LH/CG epitope replacing the residues 261–370 likely binding the blocking TRAb (B-TRAb). The capture construct was immobilized on polystyrene beads via an antibody directed against the cytosolic tail of TSHR.[12] The signal receptor was constructed from a portion of the extracellular domain (aa 21–26) of TSHR, and it was fused with secretory alkaline phosphatase (SEAP) in a buffer solution.[12] TRAb molecule had been attached to the capture receptor with one arm and to the signal receptor with the other. The amount of the TRAb bound was determined by the intensity of chemiluminescent signal developed by the reaction of SEAP with luminescent substrate.[12] We employed Immulite® for TSI evaluation, and the cut-off level suggested by the manufacturer was 0.55 IU/L while the functional sensitivity was 0.1 IU/L. TSI values exceeding 40 IU/L were reported by the laboratory as >40 IU/L. Immulite®-TSI was calibrated against 2nd IS (NIBSC 08/204).

Thyroid Scintigraphy

Fifteen minutes after intravenous administration of 185 MBq (5 mCi) of Technetium-99m (99mTc) pertechnetate, anterior images of the neck were acquired using a gamma-camera equipped with high-resolution parallel-hole collimator (Symbia®, Siemens) with an acquisition time of 10 minutes, using a 20% window centred around the 140 KeV peak of 99mTc and a 128 × 128 computer matrix. The 99mTc-pertechnetate uptake rate (TcTU) was determined using the region-of-interest technique as previously described (quantitative scintigraphy).[6] In the current retrospective analysis, images (qualitative scintigraphy) were read separately by an experienced board-certified nuclear medicine physician (AP) who was unaware of thyroid antibody results, of US results, and the final aetiological diagnosis of thyrotoxicosis/hyperthyroidism. Thyroid scans were read as diagnostic for Graves' hyperthyroidism (scintigraphy pattern 3) in the presence of diffuse thyroid overactivity with a homogeneous distribution of the tracer and reduced uptake in major salivary glands. Scintigraphy patterns 0–2 (0, normal gland uptake; 1, decreased uptake; 2, unifocal or multifocal overactive areas with reduced or suppressed uptake in the remaining thyroid tissue, consistent with autonomously functioning thyroid nodule[s] [AFTN]) corresponded to non-Graves' hyperthyroidism cases.

Colour Doppler Ultrasonography

Colour Doppler ultrasonography was performed by an experienced thyroidologist (PT) by ultrasound device (ACUSON S3000 ultrasound system, Siemens Healthineers) with a 7–14 MHz wide band linear transducer. The colour gain was adjusted, so that artefacts were prevented. The examination of thyroid vascularity (qualitative assessment) and volume, along with nodules detection and characterization, were systematically conducted. In the current retrospective study, US images were evaluated separately by PT who was unaware of thyroid antibody results, of thyroid scintigraphy results, and the final aetiological diagnosis of thyrotoxicosis/hyperthyroidism. Ultrasonography pattern 3 corresponded to 'thyroid inferno' (markedly increased colour flow Doppler signals with homogeneous distribution within the parenchyma), while echo patterns 0, 1 and 2 were defined as – normal gland echotexture and vascularity; – nodular goitre; – slightly increased colour flow Doppler of the gland, respectively. In case of disagreeing imaging interpretation, a consensus was reached.

Final Diagnosis of Thyrotoxicosis/Hyperthyroidism

To accurately evaluate the diagnostic performance of the antibody assays (TRAb and TSI assays), the colour Doppler ultrasonography and the quantitative scintigraphy (TcTU), the final diagnosis of thyrotoxicosis/hyperthyroidism was made according to the thyroid scan (qualitative scintigraphy). Therefore, for the purpose of the present study, Graves' and non-Graves' hyperthyroidism diagnosis were defined against a scintigraphy pattern 3 and scintigraphy patterns 0–2, respectively.

Statistical Analysis

For statistical purpose, laboratory values lower than functional sensitivity were considered equal to functional sensitivity (ie TSI values >40 IU/L and TRAb values >100 IU/L were considered equal to 40 IU/L and 100 IU/L, respectively). Continuous variables were described as median and interquartile range (IQR) and analysed by nonparametric test (Mann-Whitney test). Categorical variables were presented as number (percentage), and they were compared by Fisher's test. Receiver operating characteristic (ROC) curves were drawn using the final diagnosis (Graves' and non-Graves' hyperthyroidism) as reference. ROC curves for TRAb Kryptor®, Immulite® TSI assays and TcTU were analysed to select the best cut-off level and compared against one another. Clinical sensitivity and specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) were calculated against the ROC-derived cut-offs for TRAb, TSI and TcTU, respectively. Clinical effectiveness indices of echo pattern 3 for diagnosis of GD were also obtained considering the final diagnosis (Graves' and non-Graves' hyperthyroidism) as reference. Interassay correlation and agreement between TRAb Kryptor® and Immulite® TSI assays were obtained by Passing-Bablok regression and Bland-Altman test. The relationship among variables was assessed by linear regression and the Spearman's coefficient of rank correlation (r-Spearman, rs or rho). For TcTU results analysis, patients with AFTN were excluded, since diagnosis was based on scintigraphy pattern. Statistical significance was defined by a P-value <.05. Statistical analysis was performed by MedCalc software version 9.