Polyspecificity of Antimicrosomal Thyroid
Antibodies in Hepatitis C Virus–Related Infection
Katell Peoc’h, Pharm.D., Laurence Dubel, M.D., Ph.D., Olivier Chazouille `res, M.D., Ph.D., Thierry Ocwieja,
Franc ¸oise Duron, M.D., Raoul Poupon, M.D., Ph.D., and Catherine Johanet, Pharm.D., Ph.D.
Services d’Immunologie, d’He ´patologie, and d’ Endocrinologie, Hopital Saint-Antoine, AP-HP, Paris, France
OBJECTIVES: The outcome of dysthyroidism and the pres-
ence of antithyroid antibodies in patients with chronic hep-
atitis C virus (HCV) infection receiving interferon-? ther-
apy is clearly established. However, the prevalence and the
specificity of antithyroid antibodies in HCV patients before
interferon-? therapy remain controversial. The aim of the
present study is to clarify within a large population of HCV
patients the prevalence of antithyroid antibodies before in-
terferon-? therapy and to determine whether their immuno-
dominant antigen is the same as described in autoimmune
METHODS: Sera from 99 patients with chronic hepatitis C
before (n ? 99) and after (n ? 37) interferon-? treatment
were investigated for the presence of antimicrosomal and
antithyroperoxidase antibodies assessed by indirect immu-
nofluorescence and ELISA, respectively. Dot blotting on
human thyroid lysate was designed to further characterize
these autoantibodies. Data were compared to those obtained
with sera of patients with autoimmune thyroiditis (n ? 75)
and healthy subjects (n ? 96).
RESULTS: In HCV patients, antimicrosomal antibodies were
found with a higher proportion before interferon-? therapy
(12.1%) than after therapy (8%). Thyroperoxidase consti-
tutes the main antigen in only 4% before treatment, a prev-
alence similar to that observed in healthy controls.
CONCLUSIONS: The prevalence of antithyroid antibodies is
low in patients with chronic hepatitis C before interferon-?
therapy. Thyroperoxidase may not be their main target.
Further studies are required to determine whether HCV
infection leads to a breakdown of tolerance to a thyroid
self-protein other than thyroperoxidase. (Am J Gastroenterol
2001;96:2978-2983. © 2001 by Am. Coll. of Gastroenter-
First noted in 1959 (1), antimicrosomal thyroid antibodies
(TMAs) are the hallmark of autoimmune thyroiditis (AIT)
(i.e., Hashimoto’s thyroiditis and Graves’ disease). They are
usually detected by indirect immunofluorescence (IF),
ELISA, or radioimmunoassay (2). The immunodominant
antigen of these antibodies in AIT has been identified as
thyroperoxidase (3–5), a glycoprotein playing an essential
role in thyroid hormone synthesis. Antithyroperoxidase an-
tibodies (TPAs) are usually detected by ELISA, radioim-
munoassay, immunocapture, or immunoblotting (3, 6–8).
Antithyroid antibodies and dysthyroidism have also been
described in patients with chronic hepatitis C virus (HCV)
infection treated with interferon-? therapy (9–12). Further-
more, clinical manifestations of dysthyroidism are known to
be one of the most frequent side effects of this therapy in
patients with HCV-related infection (13, 14).
Interestingly, and supposedly related to the HCV-induced
autoimmunity (15–17), antithyroid antibodies have also
been reported in sera of HCV patients before interferon-?
therapy. The prevalence of antithyroid antibodies in these
patients ranged from 5% to 36% (9, 11, 18–21). In one of
these studies, surprisingly, only 50% of TMAs found in sera
of untreated HCV patients were directed to the thyroperoxi-
dase antigen (18).
These findings have raised our interest in testing the
hypothesis that TMAs are heterogeneous in HCV patients
and that HCV may induce autoantibodies directed to a
protein distinct from thyroperoxidase. Sera from patients
with HCV infection before and after interferon-? therapy
were tested for the presence of TMAs (by IF) and TPAs (by
ELISA). Data were compared to those obtained from pa-
tients with AIT and from healthy subjects. In an attempt to
further characterize the TMAs associated and/or related to
HCV, these antibodies were also assessed on human thyroid
lysate by dot blot.
PATIENTS AND METHODS
A total of 270 sera were analyzed: 99 patients with docu-
mented chronic C viral hepatitis, 75 patients with docu-
mented autoimmune thyroiditis, and 96 healthy adults.
CHRONIC C VIRAL HEPATITIS. Patients were selected
on the basis of their positivity for anti-HCV antibodies by
third generation ELISA (Pasteur, Marne la Coquette,
France) and immunoblot assay (RIBA III Ortho, Ortho
Diagnostic Systems, Raritan, NJ). All were HCV RNA
positive as assessed by hybridization using the kit Quanti-
THE AMERICAN JOURNAL OF GASTROENTEROLOGY
© 2001 by Am. Coll. of Gastroenterology
Published by Elsevier Science Inc.
Vol. 96, No. 10, 2001
plex (Chiron Diagnostics, Cergy Pontoise, France). All
had chronic elevation of serum aminotransferase activi-
ties. Eighty-nine had chronic hepatitis (90%); nine, cir-
rhosis (9%); and one, cirrhosis and hepatocellular carci-
noma (1%). All had normal thyroid-stimulating hormone
(TSH) levels and were free of clinical manifestations of
Among these patients, 99 were studied before interfer-
on-? therapy (55 males and 44 females, mean age ? 42.5
yr [range ? 27–80]), and 37 sera (26 males and 11
females, mean age ? 49.2 yr [range ? 29–66]) were
available for study after a 24-wk course of interferon-?
therapy (3 MU three times a week). None had been
treated by combination therapy with interferon alfa plus
ribavirin. Patients were sampled during the 2 months after
the end of treatment.
SEVENTY-FIVE PATIENTS WITH DOCUMENTED
AUTOIMMUNE THYROIDITIS. These patients included
70 females and five males (mean age ? 48 yr [range ?
25–76]). According to the clinical, morphological, func-
tional, and biological criteria, Hashimoto’s thyroiditis and a
Graves’ disease were diagnosed in 59 and 16 cases, respec-
tively. HCV infection was screened by third generation
ELISA (Pasteur) in all of these patients. None of these
patients were positive for HCV serological markers.
NINETY-SIX HEALTHY ADULTS. These subjects in-
cluded 60 females and 36 males (mean age ? 36 yr
[range ? 17–67]). All of these patients were HCV antibody
negative and free of clinical and biological thyroid abnor-
All sera were sampled and stored frozen at ?80°C until
DETECTION OF ANTIMICROSOMAL THYROID AN-
TIBODIES. TMAs were detected by IF on human thyroid
cryostat sections using fluorescein isothiocyanate–conju-
gated goat antihuman immunoglobulin (Bio-Rad, Marne la
Coquette, France) with a threshold at 1:10 (Fig. 1). When a
homogeneous cytoplasmic fluorescence was observed on
thyrocytes at this dilution, serial dilutions of the sera were
DETECTION OF ANTITHYROPEROXIDASE ANTI-
BODIES. TPAs were detected by ELISA on human puri-
fied thyroperoxidase (TPO Biocode, Bio-Rad; threshold ?
50 IU/L). Optical densities were read at 405 nm. Because
discrepancies between the presence of TMAs and TPAs
were observed for patients within the three groups (controls,
AIT, and HCV), their sera were double checked for TPAs
using a second ELISA kit (Autozyme TPO, BYK France, La
Me ´e sur Seine, France) with a purified recombinant thy-
roperoxidase and a threshold at 100 IU/L.
DETECTION OF ANTITHYROGLOBULIN ANTIBOD-
IES. Antithyroglobulin antibodies (TGAs) were detected
by ELISA (TG Biocode, Bio-Rad) on human purified thy-
roglobulin (threshold ? 200 IU/L).
DETERMINATION OF THE TSH LEVEL. TSH was as-
sessed by chemiluminescence assay using an automated
chemiluminescence system (Chiron Diagnostics), with nor-
mal values ranging from 0.5 to 5.5 ?IU/L.
DOT BLOTTING ON HUMAN THYROID. To eliminate
the hypothesis of false positive TMA data by IF, a dot blot
using a crude lysate of human normal thyroid was tested.
Thyroids were cut and homogenized at 4°C with an
Figure 1. Homogeneous cytoplasmic thyrocytes’ fluorescence observed with TMAs by IF on human thyroid cryostat section (?100).
AJG – October, 2001
Antithyroid Antibodies and HCV
electric crusher (Kontes, Vineland, NJ) in an NP40 lysis
buffer (Tris-HCl [pH 6.8] 60 mmol/L, NP40 1%, glycerol
10%). After a 30-min incubation on ice, the homogenate
was centrifuged (10,000 ? g at 4°C for 15 min). Supernatant
protein concentration was determined by the BCA technique
(Pierce, Rockford, IL) before freezing at ?80°C until as-
The supernatant (thyroid lysate) was prepared under both
reducing conditions (1:10 in Tris [pH 6.8] 10 mmol/L, 30%
glycerol, 3% sodium dodecyl sylfate, and 15% ?–mercapto-
ethanol, and heated 3 min at 100°C) and nonreducing con-
ditions (1:10 in Tris-HCl [pH 7.4] 10 mmol/L).
Ten micrograms of total proteins were loaded onto a
0.45-?m nitrocellulose membrane (Bio-Rad) and dried at
room temperature; nonspecific binding was blocked for 2 h
with 5% nonfat dry milk in phosphate-buffered saline
(NFDM). The membrane was then incubated overnight with
the sera diluted to 1:50 in NFDM at 4°C. After thorough
washing in 0.05% phosphate-buffered saline Tween, perox-
idase-labeled goat antihuman immunoglobulin (Bio-Rad)
was incubated (1:200 dilution in NFDM) for 1 h at room
temperature. Reactivity was detected by using 4-chlo-
STATISTICS. For the ?2test, ps ? 0.05 are considered
The prevalence of TMAs and TPAs in healthy controls was
low: 3.1% and 4.1%, respectively (Table 1). Three healthy
women displayed TMAs by IF at a low titer (average ?
1:10) associated with the presence of TPAs by ELISA
(average ? 104 IU/L for a threshold at 50 IU/L). In a fourth
case, TPAs were found alone (at 64 IU/L) (Table 2), and the
presence of TPAs for this woman was confirmed by the
Table 1. Prevalence of TMAs and TPAs in Healthy Subjects, in Patients With AIT, and in HCV Patients Before and After Interferon
(n ? 36)
(n ? 5)
(n ? 44) (n ? 55)
(n ? 11) (n ? 26)
(n ? 60)
(n ? 96)
(n ? 70)
(n ? 75)
(n ? 99)
(n ? 37)
8 8013.6 13.6 10.9 18.2
8 80 18.2
Table 2. Clinical and Biological Data for All Subjects Displaying a Discrepancy Between TMA and TPA Detection
Groups Subjects Gender Age* TMAs
(N ? 50 IU/L)
(N ? 200 IU/L) Clinical Status
HCV before interferon
Kn 6 1b
HC ? hepatocellular carcinoma; Kn ? Knodell score; ND ? not determined.
* At the time of bleeding.
† Associated cryoglobulinemia.
‡ Coinfection by HIV.
§ Concomitant presence of anti–smooth muscle antibodies.
2980Katell et al.
AJG – Vol. 96, No. 10, 2001
second ELISA. No antithyroid antibodies were found in any
of the healthy male sera.
In patients with AIT (Table 1), the prevalences of TMAs
and TPAs were 85.3% and 93.3%, respectively (ns, p ?
0.113). The average titers of TMAs and TPAs were at 1:100
(ranging from 1:10 to 1:640) and 777 IU/L (ranging from 32
to ?2000 IU/L), respectively. Six patients, including five
with Hashimoto’s and one with Graves’ diseases, were
positive for TPAs without concomitant TMAs (Table 2),
with a lower titer of TPAs (mean ? 229 IU/L) than that
observed for the other patients with AIT (mean ? 825
In untreated HCV patients (Table 1), 12 were TMA
positive (12.1%). Four (4.0%, two males and two fe-
males) displayed a positivity for TPAs (mean ? 383
IU/L, ranging from 152 to 1000). A significant difference
was found between these two prevalences (p ? 0.05),
observed independently of gender. The TMAs’ preva-
lence was higher in HCV patients before treatment when
compared to healthy subjects (p ? 0.05). The average
TMA level was at 1:10 (ranging from 1:10 to 1:20). In
four patients displaying a dual positivity, TGAs were
present. None had an abnormal TSH level or any clinical
manifestation of dysthyroidism. The clinical and biolog-
ical data for the eight remaining patients are summarized
in Table 2. The prevalence of TMAs and TPAs was
significantly lower in HCV patients before and after
treatment than in AIT patients (p ? 0.05).
For 37 patients, data were compared before and after
treatment (Table 1). No significant difference was observed
between the prevalence of TMAs and that of TPAs after
interferon therapy (Table 1). Four patients (10.8%) were
TMA positive before treatment (Table 3). One had associ-
ated TPAs, and both antibodies were still present after
interferon-? therapy. A significant increase of TSH level
was observed after the treatment for this patient. The three
patients with isolated TMAs before interferon therapy were
all negative after interferon treatment. For the two positive
patients presenting both TMAs and TPAs after treatment,
TGAs were detected at a high level with an increased TSH
level (average ? 22.2 ?IU/L).
Thirty sera were additionally tested by dot blot, includ-
ing 10 antithyroid antibody negative sera, 10 sera from
patients with AIT (displaying TMAs and TPAs), and 10
sera from patients with chronic hepatitis C (presenting
either TMAs alone or associated with TPAs). A strong
dot was obtained for all sera displaying antithyroid anti-
bodies, without any difference between sera with TMAs
alone or sera with concomitant TMAs and TPAs. A
stronger dot was obtained under reducing conditions in
all cases (Fig. 2). No dot was observed with antithyroid
antibody negative sera.
The present study has assessed the low prevalence of TPAs
and TMAs in patients with chronic hepatitis C before and
after interferon therapy. TMAs seem not to be directed only
to thyroperoxidase and are not associated with an increased
risk of clinical manifestations of dysthyroidism during in-
To validate the specificity of these two antibodies, a
population of healthy subjects was screened. As expected,
TMAs and TPAs were found with very low prevalences
(TMAs ? 3.1%, TPAs ? 4%), consistent with previous
studies (19, 22) showing that patients—especially wom-
Figure 2. Patterns observed by dot blotting on human thyroid
lysate after incubation with the following sera: 1) serum of a
healthy patient without antithyroid antibodies; 2, 3) sera of patients
with chronic hepatitis C with TMAs only; 4) serum of a patient
with chronic hepatitis C presenting both TMAs and TPAs; and 5)
sera of patient with Hashimoto’s thyroiditis presenting both TMAs
and TPAs. (A) Thyroid lysate in reducing conditions. (B) Thyroid
lysate in nonreducing conditions.
Table 3. Comparison of the Autoantibodies Detected Before and After Treatment by Interferon for the Five HCV Patients Studied
Before Interferon Treatment After Interferon Treatment
HCV (genotype 3a)
HCV (genotype 1b)
TPAs (?1000 IU/L)
TPAs (?100 IU/L)
TPAs (?300 IU/L)
F 51N 12.9HCV (genotype 1b)
TSH normal values ? 0.5–5.5 ?IU/L. ND ? not determined.
* At the time of bleeding.
AJG – October, 2001
Antithyroid Antibodies and HCV
en—may have antithyroid antibodies without clinical thy-
In patients with AIT, TMAs and TPAs were detected with
very high prevalences—85.3 and 93.3%, respectively—
consistent with previous studies. Iitika et al. (23) had indeed
shown that TMAs’ in vitro production by peripheral lym-
phocytes is correlated with TPAs. Mariotti et al. (2) have
also demonstrated that thyroperoxidase inhibited the fixa-
tion of TMAs onto thyroid microsomes.
In our study, the prevalence of TPAs in patients with
chronic hepatitis C before treatment (4%) was the same as
that observed in healthy subjects, as previously described by
Boadas et al. (19) and Colas-Linhart et al. (24) (5.2% and
Interestingly, in patients before any interferon-? therapy,
TMAs were present with a higher prevalence (12.1%) than
TPAs (4%) and than TMAs detected in healthy subjects
Similarly, in the study reported by Tran et al. (18), only
half of the patients presenting TMAs also presented TPAs.
Such a discrepancy between TPAs and TMAs had also been
described by Pateron et al. (25) in three of nine HCV
patients displaying antithyroid antibodies before treatment.
This difference of prevalence between TMAs and TPAs
supports our hypothesis that TMAs in patients with HCV
may not be solely directed to thyroperoxidase, unlike what
is observed in AIT. Such a discrepancy had already been
described in systemic lupus erythematosus patients display-
ing TMAs: only 61% of patients also presented TPAs (26).
On the other hand, previous studies have already shown a
heterogeneity of antigenic targets in HCV patients: type 2
antimitochondrial antibodies detected in HCV patients seem
not to be directed to the same antigenic target as those in
primary biliary cirrhosis (27). Ma et al. (28) and Durazzo et
al. (29) have also described the heterogeneity of liver kidney
microsomal–1 autoantibodies in autoimmune hepatitis type
2 and chronic hepatitis C sera.
In HCV patients after interferon-? therapy, antithyroid
antibodies were detected in 8% (i.e., 3/37 patients). The
prevalences of TPAs and TPAs were the same.
Therefore, comparison of our results before and after
treatment suggests that TMAs might be related to the auto-
immunity associated with or related to the C virus infection.
In addition, the lack of increased TMA prevalence (or titer)
in HCV patients during interferon therapy argues against an
underlying AIT. Only one patient developed a hypothyroid-
ism after interferon treatment associated with an increase of
TSH level. Deutsch et al. (11) had already described the
TSH level increase associated with interferon treatment.
This patient displayed both TMAs and TPAs before and
after treatment, thus supporting the idea that the presence of
TPAs may be a risk factor in developing thyroid disease.
In conclusion, our data suggest that screening patients
with chronic hepatitis C before interferon therapy for TPAs
may be useful in assessing the risk of developing a thyroid
disease during the treatment, and that antithyroid antibodies
seem to be found with a low prevalence in sera of patients
with chronic hepatitis C before interferon-? therapy. Thy-
roperoxidase seems not to represent their main target. Fur-
ther studies are in process to determine whether HCV in-
fection leads to a breakdown of tolerance to another thyroid
We gratefully acknowledge the help and cooperation of the
“Service de Me ´decine Nucle ´aire” (Pr. Askienazy, Hopital
Saint-Antoine) in TSH level determination, and the coop-
eration of the “Service de Virologie” (Pr. Petit, Hopital
Saint-Antoine) for the HCV RNA determination.
Reprint requests and correspondence: Catherine Johanet, Ser-
vice d’Immunologie, Hopital Saint-Antoine, 184, rue du Faubourg
Saint-Antoine, 75 012 Paris, France.
Received Dec. 14, 2000; accepted May 30, 2001.
1. Belyavin G, Trotter W. Investigation of thyroid antigens re-
acting with Hashimoto’s sera. Lancet 1959;i:646–52.
2. Mariotti S, Pinchera A, Vitti P, et al. Comparison of radioas-
say and haemagglutination methods for anti-thyroid microso-
mal antibodies. Clin Exp Immunol 1978;34:118–25.
3. Rapoport B, McLachlan SM. Thyroid peroxidase antibodies.
In: Peter JB, Shoenfeld Y, eds. Autoantibodies. New York:
4. Czarnocka B, Ruf J, Ferrand M, et al. Purification of the
human thyroid peroxidase and identification as the microsomal
antigen involved in autoimmune thyroid diseases. FEBS Lett
5. Portmann L, Hamada N, Heinrich G, et al. Antithyroid per-
oxidase antibody in patients with autoimmune thyroid disease:
Possible identity with antimicrosomal antibody. J Clin Endo-
crinol Metab 1985;61:1001–3.
6. Schardt CW, Mc Lachlan SM, Matheson J, et al. An enzyme-
linked immunoassay for thyroid microsomal antibodies. J Im-
munol Methods 1982;55:155–68.
7. Kaufman KD, Filetti S, Seto P, et al. Recombinant human
thyroid peroxidase generated in eukaryotic cells: A source of
specific antigen for the immunological assay of antimicroso-
mal antibodies in the sera of patients with autoimmune thyroid
disease. J Clin Endocrinol Metab 1990;70:724–8.
8. Takamatsu J, Yoshida S, Yokozawa T, et al. Correlation of
antithyroglobulin and antithyroid-peroxidase antibody profiles
with clinical and ultrasound characteristics of chronic thyroid-
itis. Thyroid 1998;8:1101–6.
9. Preziati D, La Rosa L, Covini G, et al. Autoimmunity and
thyroid function in patients with chronic active hepatitis
treated with recombinant interferon alpha-2a. Eur J Endocrinol
10. Carella C, Amato G, Biondi B, et al. Longitudinal study of
antibodies against thyroid in patients undergoing interferon-
alpha therapy for HCV chronic hepatitis. Horm Res 1995;44:
11. Deutsch M, Dourakis S, Manesis EK, et al. Thyroid abnor-
malities in chronic viral hepatitis and their relationship to
interferon alpha therapy. Hepatology 1997;26:206–10.
12. Di Bisceglie AM, Martin P, Kassianides C, et al. Recombinant
2982Katell et al.
AJG – Vol. 96, No. 10, 2001
interferon alpha therapy for chronic hepatitis C. A random- Download full-text
ized, double blind, placebo-controlled trial. N Engl J Med
13. Vial T, Bailly F, Descotes J, et al. Effets secondaires de
l’interfe ´ron alpha. Gastroenterol Clin Biol 1996;20:462–89.
14. Nagayama Y, Ohta K, Tsuruta M, et al. Exacerbation of
thyroid autoimmunity by interferon alpha treatment in patients
with chronic viral hepatitis: Our studies and review of the
literature. Endocr J 1994;41:565–72.
15. McMurray RW, Elbourne K. Hepatitis C virus infection and
autoimmunity. Semin Arthritis Rheum 1997;26:689–701.
16. Manns MP,Obermayer-Straub
autoimmunity: Mechanisms and examples in hepatology. J
Viral Hepat 1997;4:42–7.
17. Manns MP, Rambusch EG. Autoimmunity and extrahepatic
manifestations in hepatitis C virus infection. J Hepatol 1999;
18. Tran A, Quaranta JF, Benzaken S, et al. High prevalence of
thyroid autoantibodies in a prospective series of patients with
chronic hepatitis C before interferon therapy. Hepatology
19. Boadas J, Rodriguez-Espinoza J, Enriquez J, et al. Prevalence
of thyroid autoantibodies is not increased in blood donors with
hepatitis C virus infection. J Hepatol 1995;22:611–5.
20. Marazuela M, Garcia-Buey L, Gonzalez-Fernandez B, et al.
Thyroid autoimmune disorders in patients with chronic hepa-
titis C before and during interferon-alpha therapy. Clin Endo-
21. Huang MJ, Tsai SL, Huang BY, et al. Prevalence and signif-
icance of thyroid autoantibodies in patients with chronic hep-
atitis C virus infection:a prospective controlled study. Clin
22. Groves CJ, Howells RD, Williams S, et al. Primary standard-
ization for the ELISA of serum thyroperoxidase and thyro-
globulin antibodies and their prevalence in a normal Welsh
population. J Clin Lab Immunol 1990;32:147–51.
23. Iitika M, Aguayo J, Row VV, et al. Comparison of measure-
ments of in vitro production of antithyroid microsomal anti-
body versus antithyroid peroxidase antibody. Reg Immunol
24. Colas-Linhart N, Bok B, Marcellin P, et al. Antithyroid auto-
antibodies and hepatitis C virus. Hepatology 1994;20:1649.
25. Pateron D, Hartmann DJ, Duclos Valle ´e JC, et al. Latent
autoimmune thyroid disease in patients with chronic HCV
hepatitis. Hepatology 1993;17:417–9.
26. Tsai RT, Chang TC, Wang CR, et al. Thyroid peroxidase
autoantibodies and their effects on enzyme activity in patients
with systemic lupus erythematosus. Lupus 1995;4:280–5.
27. Grimbert S, Johanet C, Bendjaballah F, et al. Antimitochon-
drial antibodies in patients with chronic hepatitis C. Liver
28. Ma Y, Peakman M, Lobo-Yeo A, et al. Difference in
immune recognition of cytochrome P4502D6 by liver kid-
ney microsomal (LKM) antibody in autoimmune hepatitis
and chronic hepatitis C virus infection. Clin Exp Immunol
29. Durazzo M, Philipp T, Van Pelt FN, et al. Heterogeneity of
liver-kidney microsomal autoantibodies in chronic hepatitis C
and D virus infection. Gastroenterology 1995;108:455–62.
AJG – October, 2001
Antithyroid Antibodies and HCV