|
| |
(AT - Hashimoto's disease)
The role of genetic predisposition
It is generally well known that the most important etiological factor of
AT may be the genetic predisposition (Volpé 1985, 1986; Weetman
1991). However, the main question still to be definitively answered is
whether AT may occur only in genetically predisposed subjects. The answer
is "probably yes". The next question would be whether AT may develop in
those subjects who are not genetically predisposed. The answer in this
case would be "probably not". In both cases the probability level of both
answers is relatively high, although not absolute. The problem is that
the genetic predisposition is not yet definitively defined, which may be
apparently resolved only by molecular-biological approaches as shown in
very recent work by MacLachlan and Rapoport (1995).
The actual status of the above-mentioned answers is perhaps given by
the "Iodine and Health Statement by WHO" (1994):
"Through adaptive mechanisms, normal people exposed to excess iodine
remain euthyroid and free of goiter. In certain susceptible individuals,
iodide goiter and Hashimoto thyroiditis with hypothyroidism have been observed
after iodine intakes of 500-3000 µg/day. The prevalence of susceptible
individuals in different countries in not fully known."
The view on the important role of genetic background which is common
for Graves' and Hashimoto's diseases may be supported namely by the following
observations:
Increased incidence in females showing close interrelations between
AT and genes on the X chromosome. The female/male ratio is 4:1 to 8:1 (Volpé
1986). One of the possible mechanisms may relate to the possible effects
of estrogen on the function of T-lymphocytes.
Familial coincidence:
In this respect one of the most convincing
studies may be that by Tamai et al. (1985). They analyzed haplotypes HLA
and IgG in 243 members of 37 families and found two specific haplotypes
associated with Graves' or Hashimoto's disease. From 146 members bearing
both haplotypes for the appropriate family a total of 109 (75.5 %) had
Graves' or Hashimoto's disease, while of 97 members with only one or any
appropriate haplotype only one (1 %) was sick.
Very recently MacLachlan and Rapoport (1995) prepared recombinant thyroid
peroxidase and a spectrum of monoclonal autoantibodies against individual
epitopes of that molecule. They found evidence for the conservation as
well as inheritance of certain epitopic fingerprints in some families which
may provide insight into the genetic basis of human thyroid autoimmune
disease.
Moreover, the occurrence of Graves' disease in both siblings of dizygotic
twins is reported to be 3-9 % and that of monozygotic twins to be 30-60
% (Volpé 1985). The latter fact is strong evidence for a genetic
background of autoimmune thyroid disease, but genetic factors alone cannot
explain why some of the monozygotic siblings develop the disease while
others do not.
Association with HLA antigens:
There are several extensive reviews on this subject (Volpé 1985,
1986; Farid and Thompson 1985; Weetman 1991) showing that the association
of AT with certain HLA antigens has been repeatedly demonstrated. However,
there are apparently some clear ethnic differences, namely between intra-Caucasian
and non-Caucasian races.
Recently, methodological progress has been made in estimating HLA antigens
which has resulted in obtaining more definite informations, and even in
a new nomenclature. Thus, formerly used serological methods for HLA typing
which yielded variable results are being replaced by the methods of molecular
biology using specific DNA probes such as restriction fragments length
polymorphism (RFLP) and polymerase chain reaction (PCR).
Using such methods, Badenhoop et al. (1990) found positive DQw7 in 36
out of 64 (56 %) Hashimoto's disease patients from London and Toronto,
while it was positive in 21 out of 98 (21 %) randomly chosen controls.
Recently, Wan et al. (1995) found the association of HLA-A2 and DRB4*0101
in 69 out of 71 (97 %) of Hashimoto's disease patients from Japan compared
with 79 % in controls. In contrast, in Hashimoto's disease patients from
Japan Tamai et al. (1994) reported the frequency of DQAI*0102 of 9.5 %
and DQB1*0602 of only 2.9 % as compared with 24.9 % and 12.6 % in controls,
respectively, which shows that the HLA-DQ gene may be a genetic marker
for resistance to autoimmune thyroid disease.
Association with other autoimmune diseases:
Following the introduction
of autoantibody screening, antithyroid antibodies were found in about 6-9
% of the general population (Tunbridge et al., 1977; Hawkins et al., 1980)
and in about double that proportion in patients with organ-specific autoimmune
diseases without any clinical thyroid disorder (Betterle et al., 1984).
Actually, in 3737 subjects without clinically detectable thyroid disorders
thyroid microsomal and thyroglobulin antibodies were estimated and the
prevalence of 56/822 (7 %) and 90/1000 (9 %) was found in normal subjects
and in patients with non-autoimmune diseases of both sexes, respectively,
while 11-16 % prevalence appeared in groups of patients either with insulin-dependent
diabetes mellitus (IDDM; 137/852 = 16 %), first-degree relatives of IDDM
patients (61/470 = 13 %), vitiligo (40/373 = 11 %), alopecia areata (19/171=
11 %), and idiopathic hypoparathyroidism (4/29 = 14 %).
Recently, among 1419 children 55 cases (3.9 %) of Hashimoto's disease
were found as diagnosed by antimicrosomal and antithyroglobulin antibodies,
ultrasound, and fineneedle biopsy, which was significantly higher than
the prevalence in the normal pediatric population (Radetti et al., 1995).
Experimental autoimmune thyroiditis:
It is well known that experimental
autoimmune thyroiditis may be produced by increasing iodine intake in specific
strains of animals such as Obese Strain (OS) chicken or diabetic prone
BB/W rats (Allen et al., 1986, 1987). This topic has been recently reviewed
by Sundick et al. (1991, 1992) who concluded that such disorders may be
induced only in susceptible strains.
Possible role of increased iodine intake
Perhaps one of the first papers on the increased frequency of AT in the
US was that by McConahay et al. (1962). They collected a total of 1420
patients subjected to thyroid surgery between 1930 and 1959 and found only
one case of lymphoid infiltration in 1930, while the number of such cases
in 1959 was 185. The authors claimed that the cause of such an increased
incidence is unknown. Although this observation on the increased frequency
in consecutive patients might reflect the true increase of AT incidence,
the question of undefined territory, undefined number of population, and
undefined criteria of the patient recruiting system remained open.
Similar unclear epidemiological problems concern a series of another
observations on increased frequency of lymphoid infiltration in surgically
treated thyroid patients (reviewed by Beierwaltes, 1969). For instance,
Weaver et al. (1966) reported the data of 1992 thyroid patients subjected
to surgery between 1915 and 1963. In 1915-20 they did not find any case
of lymphocytic infiltration among 222 patients, but in 1958-63 they had
38 such cases among 240 patients.
Another repeatedly quoted paper is that by Harach et al. (1985) who
reported the histological findings in 148 cases of thyroid cancer over
20 years period in Salta (Argentina) as divided into 5 years period before
the onset of iodized salt use (8 % of cases with lymphoid infiltrate in
background thyroid tissue) and after that onset (27 to 34 % of lymphoid
infiltrates within three subsequent 5-year periods). At the same time the
urinary iodine increased considerably (from about 10 mg to about 200 mg/g
creatinine). However, it is striking that the lymphoid infiltration started
perhaps immediately after the onset of iodization and showed an approximately
fourtold increase as early as within 5 years, which differs considerably
from the slow increase of AT frequency within decades as reported by McConahay
et al. (1962), Weaver et al. (1966), Furszyfer et al. (1972) and others.
Perhaps the only one from repeatedly quoted papers on the effect of
iodine on the increased frequency of Hashimoto's thyroiditis observed some
epidemiological criteria. Thus, Furszyfer et al. (1972) used what they
called the unique epidemiological program of Rochester-Olmsted County and
within 32 years recorded 246 cases of AT. The incidence of that expressed
per 100,000 inhabitants increased from 6.5 in 1935-44 to 67 in 1955-64.
However, the total number of inhabitants of that county was not indicated.
The total number of cases divided by 32 years gives the average of about
8 cases per year. If the authors indicated the incidence of 67 cases per
year per 100,000 inhabitants for the last decade, it may be assumed that
the total number of inhabitants would be only a few thousand.
Finally, the paper by Boukis et al. (1983) also belongs to those frequently
quoted in this respect. Actually, these authors injected iodinated oil
(480,000 µg of iodine) to 58 subjects in a small Greek mountain village
on the island of Crete. After 6 months they found positive antimicrosomal
antibodies in 40 % of them. However, it should be emphasized that, first,
all subjects had goiter (44 diffuse and 14 multinodular one) before the
injection which means that their thyroid was impaired even before treatment.
Second, a megadose of iodine was administered and, third, a certain degree
of inbreeding and thus also a genetic predisposition should be assumed
when referring to a small mountain village. In contrast, Benmiloud et al.
(1994) recently administered the doses of 120, 240, 480, and 960 mg to
four groups of 40-50 schoolchildren each, respectively, and after one year
did not find antiperoxidase antibodies in any case.
Possible mechanisms by which iodide induces autoimmune thyroiditis:
From experimental studies it is well known that the iodine content in the
thyroglobulin molecule is approximately proportional to the thyroid intake
of iodine. Moreover, thyroglobulin containing from 5 to 50 or even more
atoms of iodine per molecule can be isolated from thyroids of animals subjected
to various intakes of iodine and such different iodine contents also affect
the tertiary structure of thyroglobulin (Zabel and Rawitch, 1984). In mice,
high-iodine-containing thyroglobulin elicited more anti-thyroglobulin antibody
than low iodinate thyroglobulin (Sundick et al., 1987). However, there
is strong evidence that shows that even such a high level of thyroglobulin
iodination is not any prerequisite for autoimmune thyroiditis. Thus, OS
chicken produce antibodies which react with low-iodine as well as with
high-iodine thyroglobulin. Moreover, patients with Hashimoto's thyroiditis
do not have high-iodine thyroglobulin, but rather their thyroid contains
low amount of iodine, a majority of which is inorganic and dischargeable
by KClO4 and, finally, a number of those patients have a strong antibody
response to thyroid peroxidase but not to thyroglobulin.
Another hypothesis arose from the observation by Follis in 1964 who
treated hamsters with PTU for 5 weeks to stimulate endogenous TSH, then
administered iodide after withdrawal of PTU and found acute leukocytic
infiltration of the thyroid. Mahmoud et al. (1986) expanded such observations,
showing that the iodine transported into the thyroid under such conditions
became organified and induced thyroid-cell necrosis. From those findings
a hypothesis arose that an excess of iodine under certain environmental
and genetic conditions leads to the production of reactive oxygen intermediates
and iodine radicals which can damage membrane lipids or iodinate membrane
proteins other than TG. In fact, Santisteban et al. (1985) showed in vitro
that thyroid peroxidase from Hashimoto patients may iodinate membrane tubulin
and membrane lipids. The validity of that "peroxidation hypothesis" was
later tested in OS chicken by Bagchi et al. (1990a) who showed that antioxidant
BHA (butylated hydroxyanisol) reduced the level of autoantibodies and thyroid
infiltration by leukocytes. In addition, the same authors subsequently
showed that the reduction of iodine content in the thyroid by KClO4 prevented
the autoimmune thyroiditis in similar animal model (Brown et al., 1991).
However, in conclusion it should be noted again that even the experiments
in animals did not reject the view that such mechanisms may occur only
in genetically susceptible strains.
Subconclusions I
- From the data presented above it may be concluded that, in spite of
several frequently quoted papers namely on retrospectively observed increased
frequencies of AT, the true incidence of AT has not yet been demonstrated
with the aid of epidemiological methods.
- The view that AT may develop exclusively in genetically predisposed
subjects has not yet been definitively disproved. In contrast, it has not
yet been definitively confirmed that AT can occur even in genetically notpredisposed
subjects. Thus, this question remains still open to a certain extent which,
however, is narrowing by time and growing evidence.
- The great majority of reports on the increased frequency of AT in
consecutive thyroid patients originated from the US.
However, when suggesting the coincidence of such phenomena with the
increased iodine intake, these reports actually neglected the fact that
such intake in the US has not only been increasing, but actually has been
excessive for decades. Actually, they stressed only the interrelation of
increased AT frequency with iodine intake which presumably resulted in
the myth that any, even a small, increase in iodine intake might result
in an increase in AT frequency.
|
|