Key words
TSH receptor functional antibodies - bridge immunoassay - cell-based bioassay
Introduction
Graves’ disease (GD) is an autoimmune disease caused by autoantibodies (Ab), which
bind to the thyrotropin receptor (TSHR) on the surface of thyrocytes, resulting in
uncontrolled overproduction of thyroid hormones [1]
[2]
[3]. For quantification of TSHR-Ab and confirmation of the clinical diagnosis, various
types of assay technology are commonly used in laboratory medicine. The most widely
used assays measure the competition between binding of TSHR-Ab and TSH [4]
[5] or an anti-TSHR directed human monoclonal autoantibody [6]. In contrast to these competition assays, cell-based bioassays measure increased
production of either cyclic AMP or cyclic AMP-dependent luciferase activity [7]
[8]
[9]
[10]
[11]. These bioassays exhibit high sensitivity and specificity but require experienced
lab personal.
Recently, an assay system has been reported, which directly detects the concentration
of TSHR-Ab in sera of patients by applying bridge technology [12]. It has been postulated that the bridge assay exclusively detects stimulating TSHR-Ab
(TSAb). In this assay, TSHR-Ab are detected by binding one antibody arm to a capture
receptor on the solid phase and bridging with the other arm to a detection receptor
giving a signal, thus the term bridge assay. The assay uses a mutant chimeric TSHR
(MC4) and detects TSHR-Ab based on an understanding of the structure of the extra-cellular
domain of the TSHR and its interactions with anti-TSHR-Ab [12]
[13]
[14]. In previous reports, the MC4 chimeric construct was postulated to specifically
detect TSAb [15]. These claims, however, have neither been reproduced nor verified and numerous studies
have used a MC4-expressing cell line to successfully measure and quantify blocking
TSHR-Ab (TBAb) in sera of patients with autoimmune thyroid disease [16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]. This bridge immunoassay has been clearly shown to have a high clinical sensitivity
for the detection of GD and discrimination from other thyroid diseases [12]
[20]
[21]. Up to now, however, there is neither evidence for the exclusive detection of stimulating
TSHR-Ab nor a potential differentiation of TSHR-Ab functionality when using this automated
bridge immunoassay. Therefore, the objective of this present work was to compare this
bridge immunoassay with two functional bioassays for TSHR-Ab detection and differentiation.
Patients and Methods
Patients
A total of 229 patients were included in the study. Among them 151 had GD, of those
10 patients had newly diagnosed untreated disease while 141 GD patients were on antithyroid
drugs (n=29,<6 months, n=18<12 months, and n=82>12 months, respectively, after initial
diagnosis; in 12 GD patients the time point of initial diagnosis is not known). Additional
35 patients with autoimmune Hashimoto’s thyroiditis (HT), 32 with non-autoimmune,
euthyroid nodular thyroid disease, and 11 with differentiated thyroid cancer (8 papillary
and 3 follicular) were enrolled in this study. Criteria for diagnosis of GD were based
on characteristic symptoms and signs characteristics, that is, biochemical hyperthyroidism
or endocrine orbitopathy, initially documented positivity with conventional TSHR-Ab
binding immunoassay (Cobas, Roche Diagnostics GmbH, Mannheim, Germany) [6], hypoechogenicity and increased blood flow in thyroid ultrasound and/or increased
uptake in the technetium thyroid scintigraphy. Biochemical hyperthyroidism was defined
as increased serum concentrations of free thyroxine (fT4), increased free triiodothyronine (fT3), and decreased or suppressed basal thyrotropin (TSH). Autoimmune HT was defined
as the presence of increased serum levels of anti-thyroperoxidase (TPO) Ab, eu- or
hypothyroidism and hypoechogenicity with thyroid ultrasound.
The study has been approved by the Ethical Committee of the Heinrich-Heine-University
Duesseldorf (No. 5380). All experiments have been conducted according to the principles
expressed in the Declaration of Helsinki.
Bridge immunoassay
The bridge immunoassay (Immulite 2000, Siemens) used is an automated, two-cycle, chemiluminescent
immunoassay. As described by the manufacturer, the assay employs a pair of recombinant
hTSHR constructs in a bridging immunoassay format. The capture receptor is immobilized
on the solid phase (polystyrene bead). The signal receptor is an alkaline phosphatase
labeled recombinant hTSHR in a buffer solution. In the first cycle, the sample (required
volume 50 μl) is incubated with the solid phase for 30 min, allowing the TSHR-Ab in
the sample to bind through one arm to the capture receptor. Next, centrifugal washes
remove residual sample. In the second cycle, the signal receptor is added to the reaction
tube and incubated for 30 min. The complexed TSHR-Ab bind the signal receptor through
the second arm, forming a bridge. Unbound signal receptor is then removed by centrifugal
washes. Finally, chemiluminescent substrate is added to the reaction tube and a signal
is generated in direct relation to the amount of TSHR-Ab in the sample. Incubation
cycles are 2 times 30 min. The measuring range is: 0.1–40.0 IU/l and the cut-off is
0.55 IU/l.
Bioassay for stimulating TSH-receptor antibodies
Serum TSHR stimulating antibody (TSAb) levels were measured in a blinded manner with
a FDA-cleared cell-based assay (Thyretain, Quidel, San Diego, CA, USA) according to
the manufacturer’s instructions [8]
[24]. Briefly, Chinese Hamster Ovary (CHO)-MC4 cells were seeded and grown to confluent
cell monolayers in 96-well plates for 15–18 h. Patient and control samples (required
volume 30 μl), as well as positive, reference and normal controls were diluted 1:11
in reaction buffer (RB, Quidel, San Diego, CA, USA), added to the cell monolayers,
and each plate was incubated for 3 h at 37 °C, 5% CO2. Subsequently, the CHO-MC4 cells were lysed and the relative light unit values were
quantified in a luminometer (Infinite M200; Tecan, Crailsheim, Germany). The assay
cut-off is at a percentage specimen-to-reference-ratio (SRR) of 140%. All sera were
measured in duplicates, data are expressed as mean values.
Bioassay for blocking TSH-receptor antibodies
Serum TSHR blocking antibodies (TBAb) levels were measured according to the manufacturer’s
instructions of the CE-marked cell-based assay [16]
[18]. The cut-off, initially 40% inhibition (I), was lowered in early 2017 to 34% I.
All sera (required volume 30 μl) were measured in duplicates, data are expressed as
mean values.
Statistical analysis
Prism software (PRISM 6, GraphPad Software, Inc., La Jolla, CA, USA) was used for
calculation of statistical significances and for graphical presentation. For data
showing a Gaussian distribution, the unpaired t-test was performed. For not normally distributed data, the Mann–Whitney U-test was
used. For correlation analyses, the Pearson correlation coefficient was calculated.
p-Values<0.05 were considered as significant.
Results
Demographic data
The demographic data of all study patients with the various thyroid diseases are shown
in [Table 1].
Table 1 Demograhic data of all study patients.
|
Collective
|
n
|
Age (years, range)
|
Gender
|
|
Gravesʼ disease
|
151
|
45 (18–87)
|
91% female
|
|
Hashimoto’s thyroiditis
|
35
|
43 (17–69)
|
77% female
|
|
Benign nodular goiter
|
32
|
63 (19–85)
|
84% female
|
|
Thyroid cancer
|
11
|
58 (18–72)
|
73% female
|
Age, gender and prevalence of TSHR-Ab positivity are shown in the four; different
groups with thyroid diseases.
Detailed comparison of all three TSHR-Ab assays
The 3 assays were clinically specific and negative in all patients with non-toxic
euthyroid benign nodular goiter and/or differentiated thyroid cancer. In contrast,
the bridge immunoassay and the cell-based bioassay were clinically highly sensitive
in detecting TSHR-Ab in GD patients (p<0.0001) ([Table 2]). In patients with GD, TSHR-Ab positivity was present in 127 of 151 (84.1%) and
140 (92.7%) for the bridge assay and TSAb bioassay ([Fig. 1]), respectively (p<0.001). Fifteen of 151 (10%) GD samples were TSAb bioassay positive
however bridge assay negative. Among the 151 GD patients, 125 (83%) tested concomitantly
positive in the TSAb bioassay and in the bridge assay. Both assays detected TSHR-Ab
in all ten untreated hyperthyroid GD patients ([Fig. 2]). In GD patients with a duration of less than 6 months, 27/29 (93%) and 28 (97%),
respectively, were TSHR-Ab positive with the bridge- and TSAb bioassay. In all GD
patients, the bridge assay and the TSAb bioassay correlated positively (Pearson, r=0.39,
p<0.0001) ([Fig. 3]).
Fig. 1 Comparison of the bridge assay and both bioassays for stimulating and blocking TSHR-Ab
in patients with various thyroid diseases: Prevalence of the TSH receptor antibody
positivity with the three assays in patients with Graves’ disease (GD), autoimmune
Hashimoto’s thyroiditis (HT), non-autoimmune, euthyroid nodular goiter (goiter), and
differentiated thyroid cancer. The cut-offs for the three bridge, stimulatory (TSAb),
and blocking (TBAb) bio-assays are 0.55 IU/l, 140 SRR%, and 34% inhibition, respectively.
Fig. 2 Comparison of the bridge assay and both bioassays for stimulating and blocking TSHR-Ab
in patients with Graves’s disease: Serum levels of TSHR-Ab in the 3 assays according
to the time interval since initial diagnosis of Graves’ disease.
Fig. 3 Correlation analysis of a bridge immunoassay and a bioassay for stimulating TSHR-Ab:
The bridge immunoassay and the stimulatory TSAb bioassay correlated positively (Pearson,
r=0.386, p<0.001).
Table 2 Serological data.
|
Collective
|
n
|
Bridge Assay
|
TSAb Bioassay
|
TBAb Bioassay
|
|
Gravesʼ disease*
|
151
|
127 (84%)
|
140 (93%)
|
2 (1.3%)
|
|
Hashimoto’s thyroiditis
|
35
|
2 (6%)
|
5 (14%)
|
1 (3%)
|
|
Benign nodular goiter
|
32
|
0
|
0
|
0
|
|
Thyroid cancer
|
11
|
0
|
0
|
0
|
TSAb: TSH receptor stimulating antibodies; TBAb: TSH receptor blocking antibodies;
* GD patients including those with long term disease duration.
In comparison, TSHR-Ab were present in 2 of 35 (5.7%) and 5 (14.3%) HT patients with
the bridge- and TSAb bio-assay, respectively ([Fig. 1]). One (2.9%) patient with HT and two (1.3%) patients with GD were positive in the
TBAb bioassay; these two GD patients were also positive in the bridge assay but negative
in the TSAb bioassay. The two TBAb positive samples were also positive with the COBAS
(Roche) binding assay.
Discussion
The recent guidelines of both the American [25] and European [26] Thyroid Associations for the management of Graves’ hyperthyroidism strongly recommend
the measurement of TSHR-Ab for the accurate and timely diagnosis/differential diagnosis
of GD. Indeed, compared to other diagnostic approaches, that is, thyroid scan and/or
ultrasound, TSHR-Ab measurement is more specific and less expensive demonstrating
the clinical utility and clinical relevance of these antibodies as a reliable biomarker
of the disease.
The present work reports the results of a one-to-one comparison of a recently introduced
automated “bridge” immunoassay and two FDA-cleared and/or CE-marked cell-based bioassays
for the detection of functional TSHR-Ab in patients with autoimmune and non-autoimmune
thyroid diseases. Both assays were highly sensitive for detecting TSHR-Ab in hyperthyroid
patients with untreated GD. For the bridge immunoassay, this has been shown already
in cross-sectional trials [12]
[27]
[28]. In these three cited studies, the TSHR-Ab measured with the bridge assay correlated
positively with the serum free T4 levels. The same story holds true for the clinical
sensitivity of the TSAb bioassay both in children [29] as well as in adults [20]
[21] with GD. Furthermore, on careful review of the 4 articles that evaluated the bridge
assay [12]
[27]
[28]
[30] only one study [12] looked at the interference with a monoclonal blocking TSHR-Ab and showed a positive
TSHR-Ab measurement in one of three hypothyroid, TSHR-Ab positive GD patients. These
data suggest that the bridge assay may not differentiate between stimulatory and blocking
antibodies although in the two other hypothyroid, TBAb-positive GD patients, the bridge
assay showed negative results. This was also underlined in a study of 120 consecutive
well-characterized and well-defined patients with autoimmune thyroid diseases which
included both commercially available purely-stimulating human monoclonal antibody
(M22) and blocking human monoclonal antibody (K1–70) as well as polyclonal Ab (patient
sera). This study demonstrated the variable sensitivity and analytical specificity
of seven TSHR-Ab immunoassays [20]
[21]
[23]
[31]. In contrast, other studies comparing various TSHR-Ab assays did not specifically
look at the specificity, functionality and differentiation of measured TSHR-Ab [27]
[28]
[30].
In the present study, we demonstrated a significant but low correlation between TSHR-Ab
detected in the bridge assay and TSAb measured in the bioassay and found concordant
results in 83% (125/151 GD samples) between these two assays. Since two GD patients
were TSAb-negative in the bioassay but positive in both the Immulite assay and TBAb
bioassay, the previously reported [20]
[21]
[23] unspecific detection of TBAb with the bridge assay has to be acknowledged.
To better understand any differences in the detection of TSHR-Ab between the bridge
assay and the bioassay, the TSHR constructs of both assays need to be considered.
As far as is known the MC4 molecules used in both assays are likely identical. The
manufacturer of the TSAb bioassay obtained the SG5-MC4 plasmid from the Kohn laboratory
that originally generated the MC4 construct [32]. SG5-MC4 was used then to generate a plasmid that also contained the luciferase
gene that was then used to generate the MC4-CHO-luc cell line [32]. No alteration to the MC4 sequence was made during the cloning [32]. The nucleotide sequence of the MC4 construct was determined, which allowed the
JGU lab to state exactly which amino acid (AA) residues from the rat LHCGR were inserted
and which AA residues from the human TSHR were deleted [14]. Tahara et al. [14]
[15] stated that they performed sequencing, but did not publish their sequence data of
the MC4 molecule. In comparison, the sequence of the MC4 published in Clin Exp Immunol 2010 (Figure 1 of the publication) [32] indicated several minor differences from what is indicated in Tahara et al. [14] as follows:
-
262–368 deleted from the hTSHR (rather than 262–370 as stated in [14])
-
262–334 inserted from rat LHCGR (rather than 262–329 as stated in [14])
-
S to R change at AA residue 287 of the rat LHCGR
Since the DNA sequence of the MC4 in the Tahara plasmid (SG5-MC4) was determined,
we suspect that these apparent differences were due to Tahara and co-workers inadvertently
misstating their AA sequences. Finally, Leonard Kohn is an author of the Clin Exp Immunol 2010 paper [32] and concurred with the published data. It is also important to emphasize that the
cell-based bioassay for measuring stimulating TSHR-Ab employs exactly the same MC4-construct
used in all publications from the JGU Lab.
According to the informations given by the manufacturer of the bridge assay (personal
communication with MS) this assay also uses the potentially identical MC4 molecule
for TSHR-Ab detection as described by Tahara and co-workers. As far as we know, the
automated bridge assay is also based on the identical MC4 construct which has been
described by the group of Dr. Loos and co-workers for a manual TSHR-Ab assay [12]. To the best of our knowledge there is, however, no published information available
as to whether the MC4 construct used in the bridge-assay has been sequenced or not.
Looking carefully at the published sequences of the various chimeras in the patent
of Dr. Loos (United States Patent Loos 8 999 727 April 7, 2015), the sequence data
in the patent (especially chimera B and sequence No. 2 and 4) do not show the complete
chimeric molecule (only seven amino acids of the rat LH receptor are shown; the C-terminal
part of the extracellular domain of the human TSH receptor is completely shown and
is identical to the amino acid sequence of this part of the bioassay MC4 construct).
Unfortunatley there is no reference to the 8 999 727 patent of Dr. Loos in the Immulite
package insert. Taken together, if the MC4 molecule used within the bridge assay (including
the amino acid “changes” described above) is identical to the MC4 molecule used in
the bioassay, the antibody recognition in both assays should be identical. So far,
however, no detailed information regarding MC4 construct used in the bridge assay
are available to answer this question.
Furthermore, M22 is positive in both assays [21], which is consistent with the fact that the epitope for M22 is present in the receptor
used in both assays. As previously reported [33] and confirmed [2], there is significant overlap in the binding epitopes of stimulating and blocking
TSHR-Ab on the TSHR. Therefore, it is unlikely that the chimeric MC4 TSHR is specific
for TSAb. Furthermore, the data of Tahara et al. on the TSAb specificity of MC4 have
not been reproduced. As additional evidence that MC4 is not TSAb specific, a bioassay
for measuring blocking TSHR-Ab utilizing the same MC4 cell line used in the TSAb bioassay
was developed, validated and evaluated using a human monoclonal blocking antibody
(K1–70) and human serum samples that were TSHR-Ab-positive and TSAb-negative [16]
[18]. This, now CE-marked, TBAb bioassay has performed well in identifying blocking antibodies
in patients with GD and HT.
It is unlikely that the readout could affect the results and explain discrepancies
between the bioassays and immunoassays. However, the two aforementioned blocking antibody
(TBAb)-positive patients with GD were positive in both the bridge assay as well as
in the Cobas binding assay despite major differences in design. In fact, binding assays,
generally give similar results despite differences in their design and readout.
In conclusion, the bridge immunoassay and the two bioassays are highly sensitive assays
for the detection of TSHR-Ab in patients with GD. If the chimeric TSHR molecules used
in both assays are essentially identical, then one would expect that the antibody
recognition (TSAb and TBAb) should be identical too. Data are not, however, available
to discern this with certainty. Regardless of the TSHR construct used, at the present
time the functionality of TSHR antibodies can be only confirmed with a dynamic cell-based
bioassay.
