Download Highly Sensitive lmmunoenzymometricAssay for Human Thyrotropin

dIN.
CHEM.31/4,
634-636(1985)
Highly Sensitive lmmunoenzymometricAssay for Human Thyrotropin
Yuklhiko Watanabe, Nobuyukl Amino,1 Haruo Tamaki, Mleko Aozasa, Junko Tachl, Yulchl Endo, and Kiyoshi
MIyaI
A sensitive assay procedure for immunoenzymometricassay
of serumthyrotropin(TSH) was developedby makingseveral
modifications of the Enzymun-Test#{174}
TSH kit (Boehringer,
Mannhelm GmbH). Serum samples were first incubated in
plastic tubes precoated with monoclonalantibodiesspecific
to the beta subunit of human TSH. After the tubes were
washed, the TSH bound to the tubes was detected with
peroxidase-conjugated polyclonal
antibodies to TSH. The
sensitivity of the assay was 0.2 milli-int.unit/L,and the intraand interassayCVs were <10%. Analytical
recoverywas 96
to 106%. The normalbasal rangeof TSH was 0.5 to 4.8 millimt.units/L. The basal levels of TSH in all but one of 48
thyrotoxic patients with Graves’ disease were less than 0.2
milli-int. unit/L, clearly different from those of normal subjects.
Thyrotoxicpatients in early normal pregnancy showed TSH
concentrationsof 1.7 to 2.9 milli-int.units/Lby conventional
double-antibodyradioimmunoassay,possiblyfrom cross reactivity with human choriogonadotropin,but undetectable
TSH by this method. Measurement of basal TSH by this
sensitive assay can be used as an initial screening test for
thyroid dysfunction.
Addftlonal K.yphraees: “kit” method
thyroid status
nancy
. Graves’ disease
monoclonal antibodies
.
preg-
Measurement
of serum thyrotropin
(TSH)2 is useful for
detection of various thyroid conditions, especially mild or
latent primary
hypothyroidism
(1-3), in which TSH is
increased
in serum
and thus easily detected with currently
routine RIA or enzyme immunoassay
kits. However, in patients with thyrotoxicosis, serum TSH is suppressed
by excess thyroid hormones through a negative-feedback
mechanism
(1, 4, 5). This makes it difficult to differentiate
the basal concentrations of TSH in sera of euthyroid subjects
from those of thyrotoxic patients by using commercially
available kits with low sensitivity.
Therefore, tests with
thyroliberin
(thyrotropin-releasing
hormone) have been
widely used to diagnose borderline or mild cases of thyrotoxicosis (4, 5).
In this study we developed a sensitive enzyme immunoassay for TSH in which we made several modifications of a
commercially
available kit (Enzymun-Test#{174}TSH) (6). We
also report here the clinical application of the procedure.
available
Materials and Methods
Subjects: We measured basal TSH concentrations in sera
of 100 normal subjects (69 women, 21-43 years old; 31 men,
22-35 years old) and 48 untreated
thyrotoxic patients with
Graves’ disease, including five patients in early normal
pregnancy.
Cases of subclinical
autoimmune
thyroiditis
Department of Laboratory Medicine, Osaka University Medical
School, 1-1-50 Fukushima, Fukushima-ku, Osaka 553, Japan.
‘Author to whom correspondence should be addressed.
2Nonstandard
abbreviations: TSH, thyrotropin; T4, thyroxin; T3,
T3U, T3 uptake.
Received October 29, 1984; accepted January 28, 1985.
friiodothyrornne;
634
CLINICAL CHEMISTRY, Vol. 31, No.4, 1985
were excluded from the group of normal subjects by measurements
of anti-thyroid
antibodies
(7). Graves’ disease
was diagnosed on the basis of clinical symptoms of thyrotoxicity and laboratory findings of increased thyroxin (‘F4) and
triiodothyronine
(T3) in serum, and macroaggregated
albumin T3 uptake (8).
Conventional
radioii’nmunoassay:
A commercially
available double-antibody
RIA kit (Daichi Radioisotope Laboratory, Tokyo, Japan) was used, with minor modifications. In
brief, 100 pL of serum sample or standard TSH solution was
mixed with 1004
of anti-human
TSH rabbit antibody, and
incubated for 3 h at room temperature.
Then 1004
of i251
labeled TSH was added and the mixture was incubated for
20 h at room temperature.
Finally, 1004
of anti-rabbit yglobulin goat serum was added and the mixture was incubated for 5 h at room temperature.
This was centrifuged at
2000 x g for 30 mm and the radioactivity
in the precipitate
was counted. The concentration
of standard hTSH was
calibrated with wiso standard MRC 68/38. The sensitivity of
this assay was 1.0 milli-int. unit of TSH per liter.
Immunoenzymoassay:
We used a commercially available
enzyme immunoassay
kit (Enzymun-Test#{174} TSH; Boshringer Mannheim GmbH, Mannheim, F.R.G.), with several
modifications. We incubated 2004
of serum test sample or
standard ‘NH serum (calibrated with wiio Standard MRC
68/38) with 1 mL of phosphate buffer (16 mmol/L, pH 6.9) for
17 h at 4#{176}C
(instead of 1 h at room temperature
as in the
manufacturer’s
protocol), in a plastic tube precoated with
monoclonal antibodies specific to the beta subunit of the
human TSH molecule. We washed the tube once with Tris
HC1 buffer (50 mmol/L, pH 7.4) instead of with tap water,
and we incubated
the tube with 1 mL of horseradish
peroxidase-conjugated
polyclonal antibodies against human
TSH for 2 h (instead of 1 h) at room temperature.
The
antibody-peroxidase
conjugate had been diluted 200-fold
(instead of 100-fold) with phosphate buffer (36 mmol/L, pH
6.9) 1 h before the assay.
After incubation, we rinsed the tube three times with the
Tris HC1 buffer containing 0.5 mol of NaC1 and 0.5 mL of
Tween 20 per liter, instead of rinsing once with tap water as
the manufacturer
recommended.
After adding 1 mL of
substrate solution, we incubated the mixture for 1 h at room
temperature
in the dark. For the substrate
we used ophenylenediamine
instead of the kit-supplied
ABTS#{174}
[diammonium
2,2’-azino-di-(3-ethylbenzthiazoline).6-sulfonate)]. The substrate
solution consisted of 10 mg of ophenylenediamine,
104
of H2O2 (300 milL solution), and
25 mL of citrate buffer (0.1 mol/L, pH 5.0). After this
incubation, we stopped the reaction by adding 0.2 mL of 4
mol/L sulfuric acid and measured the absorbance
of the
solution at 492 nm, in a spectrophotometer.
To measure the
lower concentrations of TSH, we prepared fresh TSH standard in sera at concentrations
of 0.2 and 0.5 milli-int. unitlL
by mixing TSH-free serum with standard TSH serum (1.0
mill-mt. unitfL) from the manufacturer.
Measurement
of thyroid hormones:
Serum T4 and T3 were
measured by radioimmunoassay
as described previously (8).
T3 uptake (T3U) was determined with a kit for measuring
macro-aggregated
albumin uptake (Amersham International, Amersham, Bucks., U.K.). The free-T4 index and free-T3
index were calculated
as T4 x T3U/mean normal T3U and T3
x T3U/mean normal T3U, respectively,
according to the
modified method of Clark and Horn (9).
7.0
Results
6.0
Figure 1 shows the standard curve for sensitive immunoenzymometric assay for human TSH. The absorbance corresponding to a TSH concentration
of 0.2 milli-int. unit/L was
significantly different (p <0.001) from that at 0 milli-int.
unitlL; thus the sensitivity
of this assay is 0.2 milli-int.
unit/L. The within-assay
CVs for samples containing TSH
at 1.0 and 15.0 mill-mt.
unitsfL were 8.0% and 5.0%,
respectively. The between-assay
CVs for TSH at 1.0 and
14.0 milli-int. unitsfL were 10.0% and 5.8%, respectively.
Analytical
recovery in the assay was 96% to 106%.
As shown in Figure 2, all but one normal subject had
detectable basal concentrations
of TSH. TSH values in
normal subjects showed a normal logarithmic distribution;
the reference interval (95% confidence limits) was 0.5-4.8
milli-int.
units/L.
All but one patient with thyrotoxic
Graves’ disease had a low (<0.2 milli-int. unit/L) basal
concentration
of TSH. Thus, by using this sensitive enzyme
immunometric
assay to measure basal TSH one can clearly
distinguish
normal subjects from thyrotoxic patients with
Graves’ disease. The one exception we saw was a patient
with Graves’ disease whose TSH was 0.9 milli-int. unit/L.
Table 1 summarizes
the results for this assay when
applied to five thyrotoxic patients with Graves’ disease who
were also pregnant.
All five patients had clinical symptoms
-J
00
0
5.0
0
8
0
E
4.0
0
0
(0
E
3.0
000
IU)
1-
0
#{149}#{149}#{149}#{149}#{149}SS#{149}#{149}#{149}#{149}S
Normal
Untreated
Subjects
Graves’ disease
Fig.2. Basal values for TSH in serum of normal
untreated patients with Graves’ disease
20A
subjects
and of
U,’int. unit
E
of thyrotoxicosis
C
hormones,
(N
a
C
.0
0.5/
double-antibody
of thyroid
radioim-
Discussion
“0
5
10
15
20
25
T S H (miUI-Int.units/L)
B
E
C
(N
a,
V
U
C
V
.0
0
concentrations
munoassay measured detectable TSH in their serum. However, the TSH beta-subunit-specific
sensitive enzyme immunoassay clearly showed that the TSH concentrations
in all
these patients were <0.2 milli-int. unitfL.
U
.0
0
and increased
but the conventional
0.05
V
.0
4
0
0.5
1.0
IS H (miHP-Int.un(tS/L)
Fig. 1. Standardcurvefor the present assay for human TSH (A); lower
concentrations of TSH are shown on a larger scale (
ISH concentrationsof 0 to1.0 mlIll4nt unit/I are the results(mean ± SD)offive
replicateassays; concentrationsof morethan 1.0 mill-mt. unit/I are the average
ofduplicate
assays
The sensitivity of an immunoassay is affected by many
factors. Among the methods used to increase the sensitivity
are the introduction
of high-affinity
antibodies,
delayed
addition of labeled antigen, prolongation
of incubation,
utilization of immunometric
assays rather than competitive
immunoassays,
and application of sensitively detectable
tracer. Delayed addition of labeled TSH and prolonged
incubation have been used to increase the sensitivity
of
inimunoassay
for TSH (10-12) but these modifications are
not suitable for use in routine tests. Very recently, Weeks et
al. (13) reported a highly sensitive immunochemiluminometric assay involving monoclonal antibody and a sensitive
chemiluminescent
tracer. More recently, a sensitive radioinununoassay
of TSH was developed by using a combination
of immunometric
assay and monoclonal antibodies (14,15).
Immunoenzymoassay
of TSH has long been used (16-18),
but one problem in this assay is sensitivity.
Recently,
Imagawa et al. (19) developed a sensitive immunoenzymoassay for TSH in which affinity-purified
polyclonal antihuman TSH IgG conjugated
with /3--galactosidase
was
used. However, this method has not been used widely as a
routine method. All these reported methods may give false
values for TSH in patients with extremely high concentrations of human choriogonadotropin, such as those in early
pregnancy.
CLINICALCHEMISTRY,Vol. 31, No.4, 1985 635
Table 1. Thyroid Hormone
T4
Values
In Serum
from Five
TI
Fr.. 14
nmol/L
MT3U,%
283
7.33
24.8
286
289
170
5.74
2.97
32.4
26.6
28.6
3.85
277
5.31
Range in normalpregnancy
90-194
2.16-3.85
Serumsamples were obtained at nine to
27.4
22-29
13
Patientswith Graves’
Fr.. T3
This work was supported by a research grant from the Intractable
Disease Division, Public Health Bureau, Ministry of Health and
Welfare; by a Grant-in-Aid for Scientific Research 56570261 (to
NA) from the Ministry of Education, Science, and Culture of
Japan; and by a grant from the Clinical Pathology Research
Foundation of Japan.
References
1. Hershman JM. Utility of the radioiminunoassay of serum
trophin in man. Ann Intern Med 74, 481-490 (1971).
CLINICAL CHEMISTRY, Vol. 31, No.4, 1985
thyro-
Disease
TSH, mIIlI-lnt.unltsfL
RIA
EIA
Indsx
Indsx
18.3
28.0
407
417
2.5
1.7
<0.2
20.1
12.7
237
247
1.4
2.7
<0.2
19.9
326
2.9
<0.2
7.5-11.5
110-188
weeks of pregnancy.MT3U,macroaggregatedalbumin 13 uptake.
The sensitivity of our iinmunoenzymometric
assay was
0.2 milli-int. unit/L, five- to 10-fold the sensitivity of conventional competitive radioimmunoassays.
This high sensitivity might be the result of use of a combination of monoclonal
antibodies specific to the beta subunit of TSH and polyclonal
antibodies and to the lowering of the nonspecific binding
background. The reproducibility of the procedure was also
satisfactory
for routine use, and patients with Graves’
disease could be clearly distinguished
from normal subjects.
Basal TSH in serum of patients with thyrotoxicosis,
measured with a highly sensitive radioassay and a chemiluminescent assay, have been reported as <0.2 milli-int. unitJL
(14, 15) and <0.05 milli-int. unitlL (13), respectively. The
reason why one of our thyrotoxic
patients had detectable
TSH (0.9 milli-int. unitJL) is not known.
Because thyroid disease is common in women of childbearing age, it is often necessary to examine thyroid function in patients who are pregnant. During pregnancy the
concentrations
of thyroxin-binding globulin, T4, and T3 in
serum are increased; the free-T4 or free-T4 index should thus
be measured to evaluate thyroid function (20). However,
normal values for free T4 differ markedly, depending on the
RIA method used (21). Furthermore,
Graves’ disease is
frequently
aggravated
in early pregnancy
(22), making
especially difficult the accurate evaluation of thyroid ftmction, particularly in early pregnancy, where the choriogonadotropin concentration is often increased to as much as
200 000 int. unitsfL (20).
As shown in Table 1, in all of our pregnant thyrotoxic
patients with Graves’ disease, serum H
was falsely
detectable by the conventional RIA, but was <0.2 milli-int.
unitfL by the modified ixnmunoenzymoassay.
This discrepancy possibly is ascribable to cross reactivity of choriogonadotropin in the conventional RIA. For assay of low concentrations of pituitary peptide hormones, monoclonal antibodies specific to the beta subunit of the peptide should be used.
During pregnancy, the thyroliberin test is not permitted for
thyroid evaluation,
and there is still much controversy
about the “normal” concentration
of free T4 in sera of
pregnant subjects. Therefore, a highly sensitive assay for
TSH should certainly be useful in these cases. Measurement
of basal TSH is the first choice as a thyroid-function
test if a
highly sensitive assay is available, as suggested by Allen
and Watson (14) and Alexander et al. (15).
636
Pregnant
<8.0
<0.2
<0.2
0.5
-2.4
2. Mayberry WE, Gharib H, Bilstad JM, Sizemore OW. Radioimmunoassay for human thyrotrophin. Ann Intern Med 74, 471-480
(1971).
3. Nelson JC, Johnson DE, Odell WD. Serum TSH levels and the
thyroidal response to TSH stimulation in patients with thyroid
disease. Ann Intern Med 76, 47-52 (1972).
4. Hershman JM. Clinical application of thyrotropin-releasing
hormone. N Engi J Med 290, 886-890 (1974).
5. Jackson IMD. Thyrotropin-releaning hormone. N Engi J Med
306, 145-155 (1982).
6. Kit instruction leaflet, Enzymun-Test5 TSH. Boehringer Mannheim Immunodiagnostics, Mannheim, F.R.G.
7. Amino N, Hagan SR, Yamada N, Refetofi’ S. Measurement of
circulating thyroid microsomal antibodies by the tanned red cell
hemagglutination technique: Its usefulness in the diagnosis of
autoimmune thyroid disease. Clin Endocrinol 5, 115-125 (1976).
8. Amino N, Yabu Y, Miki T, et al. Serum ratio of triiodothyronine
to thyroxine, and thyroxine-binding globulin and calcitonin concentrations in Graves’ disease and destruction-induced thyrotoxicosis.
J Clin Endocrinol Metab 53, 113-116 (1981).
9. Clark F, Horn DB. Assessment of thyroid function by the
combined use of the serum protein-bound iodine and resin uptake of
‘311-triiodothyronine. J Clin Endocrinol Metab 25, 39-45 (1965).
10. Hall R, Amos J, Ormston B,J. Radioimmunoassay of human
serum thyrotrophin. Br Med J i, 582-585 (1971).
11. Pekary AE, Hershman HM, Parlow AF. A sensitive and precise
radioimmunoassay for human thyroid-stimulation hormone. J Clin
Endccrinol Metal, 41, 676-684 (1976).
12. Wide L, Dahlberg PA. Quality requirements of basal s-TEH
assays in predictingan s-TSH response to TEN. Scand J Clin Lab
Invest 40, 101-110 (1980).
13. Weeks I, Sturgess M, Siddle K, et al. A high sensitivity
immunochemiluminometric assay for human thyrotrophin. Clin
Endocrinol 20, 489-495 (1984).
14. Allen KR, Watson D. Thyrotropin as the initial screening test
for thyroid disease. Clin Chem 30, 502-503 (1984).
15. Alexander WD, Kerr DJ, Ferguson MM. First-line test of
thyroid function. Lancet ii, 647 (1984). Letter.
16. Miyai K, Ishibashi K, Kumahara Y. Enzyme-linked immunoassay of thyrotropin. Clin Chim Acta 67, 263-268 (1976).
17. Kate N, Naruse H, Irie M, Tsuji A. Fluorophotometric enzyme
immunoassay of thyroid-stimulating hormone. Anal Biochem 96,
419-425 (1979).
18. Miyai K, Ishibashi K, Kawashima M. Two-site inununoenzymometric assay for thyrotropin in dried blood samples on fflter
paper. Clin Chem 27, 1421-1423 (1981).
19. Imagawa M, Ishikawa E, Yoshitake S, et al. A sensitive and
specific sandwich enzyme immunoassay for human thyroid-stimulating hormone. Clin Chim Ada 126, 227-236 (1982).
20. Amino N, Yamada T, Mitsuma T, et al. Increase in plasma
thyrotropin-releasing hormone in normal human pregnancy. J Clin
Endocrinol Metab 53, 1288-1290 (1981).
21. Amino N, Nishi K, Nakatani K, et al. Effect of albumin
concentration on the assay of serum free thyroxin by equilibrium
radioimmunoassay with labeled thyroxin analog (Amerlex5 Free
T4). Clin Chem 29, 321-325 (1983).
22. Amino N, Tanizawa 0, Mon H, et al. Aggravation of thyrotoxicoals in early pregnancy and after delivery in Graves’ disease. J
Clin Endocrinol Metab 55, 108-112 (1982).