Key words
Graves` disease - TRAb - TSI
Introduction
Graves´ disease (GD) is an autoimmune disease accompanied by the production
of thyroid stimulating hormone (TSH)-receptor antibodies (TRAb) that bind to the
TSH-receptor (TSH-R) and activate the cyclic adenosine monophosphate (cAMP) signal
transduction pathway [1]
[2]. TSH-receptor-stimulating
immunoglobulin (TSI) is a subtype of TRAb that causes the stimulation of thyroid
hormone production [3]
[4]. Other types of antibodies block the
action of TSH-R and TSH-R-binding inhibitory immunoglobulins, yet others are neutral
without any functional effect [5]
[6]. Graves´ orbitopathy (GO) is
clinically relevant in approximately 50% of patients with GD, with severe
forms affecting 3–5% of patients with GD [7]. The risk of GO has been linked to TSH
receptor stimulation [8]
[9]
[10], high TRAb levels [11],
smoking [12], low levels of thyroid
peroxidase antibodies [13]
[14]
[15]
[16]
[17], and thyroglobulin antibodies [14]
[17]. Measurement of TRAb is a standard method for confirmation of
diagnosis, monitoring of therapy, and prediction of remission or relapse in patients
with GD [18]
[19]
[20]
[21]. A potential pitfall
of using TRAb is that the method will also measure blocking and neutral antibodies,
if present.
At present, there are two methods for analyzing antibodies against TSH-R. The
third-generation thyrotropin-binding inhibitor immunoglobulin (TBII) assay and the
functional thyroid-stimulating immunoglobulin bioassay. The TBII assay uses the
ability of TRAb to inhibit the binding of radiolabeled TSH to TSH receptors. In
contrast, the bioassay measures cAMP production after TSH-R antibodies bind to the
TSH receptor. This method makes it possible to identify the functional properties of
TSH-R antibodies.
The IMMULITE 2000/2000 XPi TSI assay uses recombinant thyroid stimulating
hormone receptor (TSHR) chimeras constructed from the N-terminal domain of the TSHR,
which contains the TSI epitopes, and the luteinizing hormone (LH) receptor from
which the blocking epitopes are absent and used as capture antigens. This assay is
suggested to be more specific for activating antibodies that cause hyperthyroidism
and is less influenced by blocking antibodies [22]. However, there is a lack of clinical studies that demonstrate a
clear benefit of using TSI versus TRAb.
Tozzoli et al. examined a cohort of 72 patients with untreated GD. The diagnostic
performance of the Immulite TSI assay in GD patients was comparable to that of
current TRAb assays [23]. In a study
including 166 patients with GD, the TSI method showed perfect sensitivity according
to the clinical diagnoses of GD [24].
GO is caused by the stimulation of TSH-R in the orbita [9]. Severe forms of GO may cause
irreversible eye complications and, in some severe cases, can even threaten vision
[25]. A delay in the diagnosis is
associated with a worse prognosis of GO. Therefore, there is an increased need for a
scoring system that includes different markers for the prediction and assessment of
GO.
In a retrospective study from Korea, a total of 112 patients were investigated
whether serum TSH-R antibodies in newly diagnosed, untreated GO patients were
predictive of disease course beyond 1 year after initial GO diagnosis. The
predictive power of the third-generation TBII assay and Thyretain TSI Stimulating
Reporter BioAssay were similarly strong [26].
In a recent prospective study, we investigated a cohort of 30 patients with newly
diagnosed GD and presented data that may suggest that evaluation of thyroglobulin
levels in GD could contribute to identifying patients at increased risk of
developing GO [27]. The TRAb levels were
not significantly correlated to GO in that study. To investigate whether TSI is a
more sensitive biomarker than TRAb for the evaluation of GO, we have now analyzed
TSI with the IMMULITE 2000 Systems Analyzers using stored blood samples from the
same cohort of 30 patients that participated in the previously reported study.
Materials and Methods
Study subjects
Thirty patients with de novo GD were recruited at the Uppsala University
Hospital. GD diagnosis was based on decreased levels of TSH and sero-positivity
for TRAbs. In one patient, the TRAb levels were below the reference range (1.7
IE/L, reference<1.75). This patient otherwise had symptoms,
laboratory findings, and homogenously increased uptake on thyroid scintigraphy
in line with GD. The median age of the study cohort was 55 years (range
35–72 years). Two were smokers and 29 were women.
The study consisted of six visits: at baseline, 6 weeks, 12 weeks, 6 months, 12
months, and 24 months from diagnosis. At the first visit (baseline), all
patients underwent an examination, including a recording of demographic
characteristics (sex and age) and medical and family history. At each visit,
blood was sampled to measure TSH, fT4, fT3, TRAb, and TSI.
All 30 patients received the conventional block and replace treatment (tiamazol
and levothyroxine). Three patients underwent total thyroidectomy, and five
received radioiodine treatment during the study period. These patients were not
included in the longitudinal analyses described below. Out of the remaining 22
subjects, 21 received antithyroid drug (ATD) for 18–24 months until
negative TRAb. One patient had a spontaneous recovery before treatment
initiation but was included in the longitudinal analyses. GO was identified by
clinical signs and symptoms, and the severity of GO was assessed according to
The European Group on Graves´ Orbitopathy guidelines [28]. All patients were examined by a
doctor and a nurse at each visit, and the patients received clear information
about eye symptoms. GO was classified as ‘mild’ with symptoms
such as gritty sensation and tearing due to dry eyes, caruncle swelling
and/or redness, upper eyelid retraction, or as ‘moderate to
severe’ in the instance of redness and/or swelling of the
eyelids, chemosis, pressure or pain in the eyes, exophthalmos, diplopia or signs
of optic nerve compression. The eye signs and symptoms were detailed in the
medical records by the attending endocrinologist as well as by the research
nurse at every visit [27].
Assays
Plasma TSH (reference interval 0.4–4.0 mIU/L), free T4 (reference
interval 12–22 pmol/L), free T3 (reference interval
3.1–6.8 pmol/L), TRAb (reference<1.75 IE/L),
were all measured on Cobas immunoanalyzers (e 801, Roche Diagnostics, Rotkreuz,
Switzerland) at the Department of Clinical Chemistry of the Uppsala University
Hospital. The laboratory is accredited according to SS-EN ISO 15189:2012 by
Swedac (Borås, Sweden). The total coefficient of variation (CV) for the
methods was TSH: 5% at 0.1 mIU/L, free T4: 4% at 16
pmol/L, free T3: 4% at 15 pmol/L, Tg: 4% at
23 µg/L, and TRAb: 6% at 17.4 IE/L. TSI
assay was conducted on the IMMULITE 2000 and IMMULITE 2000 XPi Immunoassay
Systems (Siemens Healthineers, Erlangen, Germany). The total CV was 5%
at 0.7 IU/L. All laboratory testing was performed blinded to clinical
information.
Statistical analysis
All statistical analyses were performed using R 4.0.2. Data are presented as
median (range) unless otherwise indicated. P
values+<+0.05 were considered significant.
Spearman’s Rank correlations between TSI, TRAb, and fT3 were performed
at baseline and every follow-up visit. At v1, TSI and TRAb values for GO
(n=17) and non-GO (n=13) patients were tested for the
significant difference using the Mann-Whitney-U test.
For all subjects who received ATD and did not undergo surgery or radioiodine
treatment (n+=+22), the area under the curves
(AUC) of TSI and TRAb after treatment (v2–6) were calculated with the
trapezoid method and compared between GO (n=12) and non-GO
(n=10) with Mann-Whitney-U tests.
Results
For all subjects, fT3 but not fT4 correlated to both TSI
(r
s=0.44, p=0.017, [Fig. 1a]) and TRAb
(r
s
=0.41 p=0.025; [Fig. 1b]) at baseline (n=30).
During the remainder of the visits (n=22), no significant correlations were
observed between fT3 and TSI or TRAb (data not shown), whereas fT4 correlated to
TRAb at v6 only (r
s=0.53, p=0.012) and did not
correlate to TSI at any visit. TSI was highly correlated with TRAb at baseline
(r
s
=0.64, p<0.001; [Fig 1c]) and across all visits and
treatments (r
s
=0.60, p<0.001 for all
visits, data not shown). There were no correlations between the levels of TSI or
TRAb in relation to the Tg levels (data not shown).
Fig. 1 Correlation between fT3 vs. TSI (a), fT3 vs. TRAb
(b), and TSI vs. TRAb (c) at baseline. Missing TSI and
TRAb value for one subject.
r
s=Spearman’s Rho coefficient. Grey
areas in A and B represent the normal range of fT3. TSI: thyroid stimulating
immunoglobulins; TRAb: TSH-receptor antibodies.
In total, 17 patients were diagnosed with GO at baseline (n=11) or during the
study follow-up period (n=6). TSI and TRAb did not differ significantly
between GO and non-GO patients at baseline (GO:non-GO=17:13), while fT3 and
fT4 were higher in the GO group than the non-GO group at baseline ([Table 1]). In the ATD group
(n=22), TSI levels appeared to be higher in GO (n=12) vs. non-GO
(n=10) during the first three visits, but this difference was not
significant when comparing the AUC during the whole follow-up period
(p=0.49) ([Fig. 2a]), and TRAb
levels ([Fig. 2b]) were similarly
reduced and normalized in both groups.
Fig. 2 TSI and TRAb levels presented as median and range in all visits
for the ATD group (n+=+22) divided into GO
(n+=+12) and non-GO
(n+=+10), showing a higher but not
significant course for TSI during the first 3 months (a) in GO vs.
non-GO group, whereas TRAb levels (b) were similarly reduced a ATD:
antithyroid drug; GO: Graves´ orbitopathy; TSI: thyroid stimulating
immunoglobulins; TRAb: TSH-receptor antibodies.
Table 1 TSI, TRAb, fT3, and fT4 at baseline presented as
median and range in GO (during the whole study) vs non-GO, compared
using Mann-Whitney-U-test.
|
GO (n=17)
|
Non-GO (n=13)
|
p-value
|
|
TSI
|
4.5 (1–40)
|
2.6 (0.7–19)
|
0.535
|
|
TRAb
|
6.3 (2.7–78)
|
4.3 (1.7–29)
|
0.053
|
|
fT3
|
18.3 (9–34)
|
12.4 (5.6–28)
|
0.026
|
|
fT4
|
48 (25–100)
|
35 (17.6–60)
|
0.008
|
GO: Graves´ orbitopathy; TSI: thyroid stimulating immunoglobulins;
TRAb: TSH-receptor antibodies; fT3: Free triiodothyronine; fT4: free
thyroxine.
Discussion
GO is a common complication that is clinically relevant in about 50% of
patients with GD. The symptoms are often mild, but severe forms of GO occur in
3–5% of patients [7]. GO
may begin early in the course of the disease, but it is not uncommon that GO occurs
later, during the follow-up period, despite ongoing treatment with ATD [9]. There are difficulties in identifying
patients with a risk of developing GO and those who will develop severe forms of eye
complications. There is, therefore, a need for better markers that can be useful for
the management of GO. Previous studies have suggested the benefits of using TSH
receptor-stimulating antibodies for the management of GO [29]. In this study, we made a comparison
between TSI and TRAb in the management of GO but could not detect any superiority of
TSI compared to TRAb in the prediction of GO.
TSH-R is well-documented to be expressed in the orbita [30]. Stimulation of TSH-R in the orbita
leads to stimulation of fibroblasts resulting in fibroblast proliferation, an
increase in hyaluronic acid production, and an increased inflammatory response [9]. Higher levels of TRAb have been
associated with an increase in the prevalence of GO [13], which we also noted in the present
study. A Swedish report found that a TRAb level of 6.3 IU/L at diagnosis of
GD was associated with an elevated risk of developing GO [16]. However, current standard methods,
which are widely routinely used, measure all antibodies, i. e., if present,
also blocking and neutralizing antibodies. Techniques that measure only the
stimulating antibodies would possibly increase the precision of the diagnosis and
follow-up of patients with GD, including those suffering from eye complications.
Currently, only bioassays appear to be able to detect purely stimulating antibodies.
Thyretain by Quidel is one such method that uses engineered CHO cells with chimeric
TSH-R (MC4) to measure luciferase gene expression due to activation of the cAMP
signal transduction pathway [5]. Lytton
et al. showed that the clinical sensitivity and specificity of the Mc4/TSI
were greater than TRAb in GO [5]. An
interesting new technique, “aequorin TSAb assay,” has been recently
developed [31]; it uses a cell line CHO
transfected with human TSH receptor, cyclic nucleotide-directed calcium channel, and
aequorin. These bioassays appear to be precise and might add to the prediction of
the course of disease and risk of GO, but it is unlikely that they will be developed
into clinical routine. However, the TBII assay has satisfactory clinical sensitivity
and specificity (97.4% and 99.2%, respectively). In this setting,
the bioassays should be reserved for fine and complex diagnoses and for particular
clinical conditions [32]
[33].
The IMMULITE 2000 TSI assay by Siemens is a bridge-binding immunoassay by which TSH-R
antibodies bind to a pair of recombinant TSH-R constructs; one is a capture, and the
other is signal receptor for measuring secreted alkaline phosphatase. This method is
supposed to provide a more sensitive measurement of stimulating antibodies compared
to the routinely used competitive-binding assay. In this prospective study of
patients with newly diagnosed GD, we could, however, not show any significant
difference between TSI and TRAb, neither among GD patients at diagnosis or follow-up
nor between those with and without GO. The correlation between fT3 and TSI and fT3
and TRAb in this study is statistically significant but with low coefficients. This
may be related to the bioactivity of the antibodies and their effect on the degree
of hyperthyroidism. Typically, high antibody titers lead to more hyperthyroidism,
but pronounced hyperthyroidism may also occur at lower antibody levels and vice
versa. The correlation between TSI and TRAb at baseline and in the subsequent visits
was good throughout, indicating that TSI is not more specific than TRAb in assessing
the severity of hyperthyroidism.
Our results are in line with a newly published study from Australia [34], in which 140 participants were
recruited, of which 75 (53.6%) had GO. Although Immulite TSI level was
higher in the presence of GO, it showed poor diagnostic accuracy and no correlation
with clinical markers of GO severity or activity. To our knowledge, only one more
study has been reported to investigate TSI in GO [35]. In that study, TSH-R antibodies were
measured with Immulite TSI assay and with Elecsys IMA in patients with moderate to
severe GO. Both methods showed a comparable correlation with GO activity during
ivGCs therapy. The data from our prospective study confirm and add to the findings
from these studies showing similar performance of the two assays in clinical
routine.
The weakness of our study is the relatively small sample size. However, the
prospective design with continuous follow-up of the clinical and laboratory status
for two years gives the present study sufficient statistical strength to rule out
significant differences regarding the parameters we have measured.
In conclusion, TSI is highly correlated with TRAb in patients with GD. The tests did
not differ significantly between GO and non-GO patients before the treatment of GD.
After treatment, both TSI and TRAb levels were reduced and normalized in both
groups. The present study does not support any added benefit of TSI compared to TRAb
for the prediction and management of GO.
Patient Consent
Consent was obtained from each patient after a full explanation of the purpose and
nature of all procedures used.
Ethical Approval
The study complied with the Declaration of Helsinki and was approved by the Regional
Ethics Committee in Uppsala (Dnr 2015/469).
Author Contributions
All authors contributed to the study conception and design. Material preparation,
data collection, and analysis were performed by SK, ML, and ÖL. The first
draft of the manuscript was written by SK and ÖL, and all authors commented
on previous versions of the manuscript. All authors read and approved the final
manuscript.
Funding Information
Uppsala Universitet —
http://dx.doi.org/10.13039/501100007051;
