Comparison of a Bridge Immunoassay with Two Bioassays
for Thyrotropin Receptor Antibody Detection and Differentiation
Authors’ response
The problem with these criticisms is that the intended use of the Thyretain TSI
Reporter BioAssay and any statement in the package insert apply to the use of
the Bioassay during the course of patient management in a clinical setting. The
purpose of this study [2] and that of
numerous other published studies [3]
[4]
[5] is to evaluate various TSH-R antibody
assays to better understand their performance and clinical utility. This is a
research study and the results are not being used to manage individual patients
but rather to better understand how results from different TSH-R Ab assays could
be interpreted. The patient population used for the study was well-described in
Table 1 in reference [2].
In their second criticism, Drs. Kiaei and Molinaro state that our analysis of the
data is inaccurate. They state that we fail to mention or consider, the fact
that the Thyretain TSI Reporter BioAssay reports net stimulating activity as
documented in previous publications by some of the authors of this study.
Although blocking antibodies do not generate a signal in the Thyretain TSI
Reporter BioAssay, blocking antibodies, when present, interfere with that
assay’s measurement of stimulating antibodies. As such, the assay
reports “zero” net stimulating activity for samples containing
equal activity of stimulating and blocking antibodies. Therefore, it is possible
that stimulating antibodies were present in the patient samples despite the
negative result reported by the Thyretain TSI Reporter BioAssay.
Authors’ response
The presence of anti-TSH-R blocking antibodies in patients with autoimmune
thyroid disease is a well-established phenomenon supported by a large body of
literature [6]
[7]. The tertiary referral thyroid lab at
the Johannes Gutenberg University (JGU) Medical Center, Mainz, Germany has
considerable experience in testing for blocking antibodies [8]
[9]
[10]
[11]. The JGU lab has observed the
presence of blocking TSHR-Ab both prior to starting an ATD treatment as well as
during the medical management of the thyroid dysfunction. The CE-marked
Thyretain TBI bioassay has been well-characterized and has performed well in
identifying blocking antibodies in patients with GD and HT [9]
[10].
Despite the hypothetical possibility that a patient could have both stimulating
and blocking antibodies that could cancel each other out in bioassays, there is
no direct experimental evidence of such an effect when testing patient serum. It
is true that the Thyretain TSI Reporter BioAssay measures net stimulatory
activity and this is, in fact, physiologically relevant because the thyroid
gland responds to net stimulatory activity. In this context, it is correctly
stated by the authors of the letter that small amounts of a monoclonal TSAb
(M22) in the presence of higher amounts of a monoclonal TBAb (K1-70) result in a
negative result in the Thyretain TSI Reporter BioAssay [12]. In our present study, the samples that
were negative in the Thyretain TSI Reporter BioAssay, but positive in the
blocking bioassay, had net blocking activity. The fact that these samples were
positive in the IMMULITE® 2000/2000 XPi TSI (bridge
immunoassay) is interpreted by Drs. Kiaei and Molinaro as being TSI positive
because the bridge immunoassay is purported to be TSI specific. There is,
however, no evidence that the bridge immunoassay is specific for stimulating
TSH-R Ab. The bridge immunoassay is a novel binding assay designed to detect
bridging of two TSH-R fragments following binding by an anti-TSH-R antibody. It
was designed using a mutated form of the TSH-R referred to as MC4. This MC4
molecule was initially thought to bind only to stimulating antibodies. There is,
however, significant published evidence that demonstrates that blocking
antibodies bind to the MC4, and, in fact, the MC4 molecule is used in the
well-characterized Thyretain TBI bioassay that detects blocking antibodies [8]
[11]. The sequence of the MC4 molecule
originally constructed by the Kohn group (GenBank Accession number M63925) and
used for the Thyretain TSI Reporter BioAssay was published [13]. There is, however, uncertainty
regarding the sequence of the MC4 molecules used in the IMMULITE
2000/2000 XPi TSI assay and it would be useful for it to be made public.
It is noteworthy that in the 510(k) submission to the FDA (K152061, July 24,
2015), it is stated that the IMMULITE 2000/2000 XPi TSI
assay...”employs two recombinant chimeric human TSH receptors (hTSHR)
where the major epitope for the blocking antibody is replaced.” It is
not clear if this represents two different ‘MC4’ molecules and
whether this could influence anti-TSH-R antibody recognition. Regardless, since
the TSAb and TBAb antibody recognition sites have not been fully defined, there
is no evidence that supports the notion that the major epitope for blocking
antibody could be replaced as claimed.
Furthermore, there is experimental evidence that the IMMULITE 2000/2000
XPi TSI assay is not TSI specific. For example, a human monoclonal TSH-R Ab
(K1-70) that is purely blocking is positive in the bridge immunoassay [14]. In addition, a number of serum samples
that are positive in a blocking bioassay (i. e. have net blocking
activity) have tested positive in the IMMULITE 2000/2000 XPi TSI assay
as well as other anti-TSH-R binding assays [12]
[15]. These empirical data lead us to
conclude that it is far more likely that the IMMULITE 2000/2000 XPi TSI
assay is not specific for TSI rather than that the TSI bioassay results were
false negative. In our opinion, to be able to interpret that samples that are
positive in the IMMULITE 2000/2000 XPi TSI assay and positive in a
blocking bioassay as positive for both TSI and TBI requires experimental
evidence that the IMMULITE 2000/2000 XPi TSI assay is actually specific
for TSI. To our knowledge, such evidence has not been published. We conclude,
therefore, that the IMMULITE 2000/2000 XPi TSI is most likely not
specific for the detection of TSAb.
Another concern that has been raised is that testing for blocking antibodies in
the presence of stimulatory antibodies may give false positive results due to
the possibility that TSI may have partial antagonist activity which may be
interpreted as blocking activity. This hypothetical possibility has never been
demonstrated directly. In addition, in our experience, almost all patients who
are positive for TSI test negative in the blocking bioassay, which suggests that
partial antagonist activity of TSI is rare and/or that the blocking
bioassay, as designed, does not detect partial antagonist activity.
In conclusion, although we cannot exclude that TSAb and TBAb might co-exist in
our two initially hypothyroid patients, there is no evidence to support the idea
that the TSAb antibodies are specifically recognized by the IMMULITE
2000/2000 XPi TSI assay. Clinically more relevant are, however, the TBAb
present in these patients’ serum which, in our opinion, are also being
detected in the IMMULITE 2000/2000 XPi TSI assay.
