| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Special Articles |
Department of Bacteriology and Immunology, University of Helsinki (M.G., A.M.); Departments of Internal Medicine (T.T.) and Immunology (S.K., A.M.), HUCH Laboratory Diagnostics; Hospital for Children and Adolescents, Helsinki University Central Hospital (P.B., M.K., J.P.); National Public Health Institute (P.B.); and Finnish Red Cross Blood Transfusion Service (J.P.), FIN-00014 Helsinki, Finland; and GKT School of Medicine (M.R.C.), London SE5 9PJ, United Kingdom
Address all correspondence and requests for reprints to: Aaro Miettinen, M.D., Ph.D., Department of Bacteriology and Immunology, Haartman Institute, P.O. Box 21, Haartmaninkatu 3, University of Helsinki, FIN-00014 Helsinki, Finland. E-mail: aaro.miettinen{at}helsinki.fi
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
APECED is rare, but it is relatively enriched among the Finns (incidence, 1:25,000) (9), the Sardinians (10), and the Iranian Jews (11). The phenotype and the age at manifestation of the disease components vary widely between different individuals, probably due to differences in genetic background and environmental factors. In Finland the disease usually manifests in early childhood, and more than two thirds of the patients develop chronic mucocutaneous candidiasis, hypoparathyroidism, and adrenocortical failure (9). In addition, gonadal failure, chronic autoimmune hepatitis, pernicious anemia, and ectodermal dystrophies are common (12). Insulin-dependent diabetes mellitus (type 1 diabetes) is also frequent. In our previous report on type 1 diabetes in the Finnish patients with APECED, 8 (17%) of the 47 patients were affected (13). As a sign of autoimmunity the patients have autoantibodies against affected organs (9, 14). In contrast to adrenal antibodies (ab), which have a high positive predictive value (92%) for the development of adrenal failure (15), islet cell ab (ICA) and GAD65 ab have a positive predictive value of only 27% for type 1 diabetes (13). Thus, better markers for active ß-cell destruction are needed. Antibodies against a tyrosine phosphatase-like transmembrane protein, IA-2, have been reported to be more specific markers than GAD65 ab or ICA for the development of type 1 diabetes among first degree relatives of patients with type I diabetes (16, 17) and in patients with autoimmune syndromes (18, 19, 20). The combined use of GAD65 ab, IA-2 ab, and IAA may give the best results. Whether this is also true for patients with APECED is not known. We have now studied a unique material of 60 well characterized Finnish patients with APECED, including 12 diabetic subjects, to evaluate the predictive values of ß-cell antibodies for the development of type 1 diabetes in this disease. In addition, we have evaluated HLA II haplotypes for association with diabetes in these patients.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The criteria for the diagnosis of APECED were the presence of
two of the following: hypoparathyroidism, adrenocortical failure, and
chronic or recurring mucocutaneous candidiasis (or one of these if a
sibling had three) (9). In this study 68 Finnish patients
with APECED were included. HLA genotype data were available from 67
patients. Serum samples taken between January 1972 and December 1998
were available from 60 patients with APECED, including samples from 12
patients who developed type I diabetes 1.3–23.1 yr after the first
serum collection (Table 1
). Aliquoted
serum samples were stored at -20 C before testing. Of the Finnish
patients, 100% had had chronic candidiasis, 85% hypoparathyroidism,
72% adrenocortical failure, and 18% type 1 diabetes. All of these
patients were examined at the Hospital for Children and Adolescents,
Helsinki University. The ethical committee of the hospital approved the
studies.
|
An iv glucose tolerance test was performed by a standard method as described previously (13). Serum insulin and C peptide concentrations were measured by RIA techniques [Pharmacia (Uppsala, Sweden) and BVK-Mallinkrodt (Germany), respectively]. The first phase insulin response (FPIR) to glucose was defined as both the absolute and the incremental sum of insulin concentrations at 1 and 3 min (13).
IA-2 and GAD65 antibodies
The last available serum sample from each patient was tested for each anti-ß-cell ab. Subsequently, all available sera from the patients with type 1 diabetes or positive for autoantibodies were tested. Thus, for each patient we tested one to seven samples taken at 0.5- to 7.0-yr intervals. IA-2 ab were detected as described for GAD65 ab (21). The IA-2ic construct used included amino acid residues 603–979 of human IA-2 (22). Briefly, IA-2ic complementary DNA in pSP64 polyadenylase (A) (Promega Corp., Madison, WI) was in vitro transcribed and translated with SP6 ribonucleic acid polymerase in a TNT-coupled rabbit reticulocyte lysate (Promega Corp., Madison, WI) in the presence of [35S]methionine (1000 Ci/mmol; NEN Life Science Products, Boston, MA). For immunoprecipitation, serum samples diluted 1:40 in precipitation buffer (150 mmol/L NaCl, 20 mmol/L Tris-HCl, 0.15% Tween-20, 0.10% BSA, and 0.02% sodium azide) were incubated on a shaker at 4 C overnight with the 35S-labeled IA-2ic. Then, 50 µL 50% protein A Sepharose CL-4B (Pharmacia Biotech, Uppsala, Sweden) in the precipitation buffer was added to isolate the immune complexes. After incubation on a shaker at 4 C for 1 h, the samples were washed four times in the precipitation buffer, scintillation fluid (OptiPhase Supermix, Wallac, Inc. Turku, Finland) was added, and the activity of the precipitates was measured in a liquid scintillation counter (1450 Microbeta Trilux, Wallac, Inc.). All samples were tested in triplicate. The results are expressed in relative units (RU), calculated as RU = serum sample cpm - negative control serum cpm/positive control serum cpm - negative control serum cpm. Strongly positive sera (>60 RU) were further diluted 1:10, and the RU at this dilution was multiplied by 10 for the final RU. The cut-off point for this assay (2.5 RU) was established at the 99th percentile for 55 Finnish nondiabetic children and adolescents and 45 healthy adult blood donors. The sensitivity of our assay was 67.1% when sera from 73 children with newly diagnosed type 1 diabetes (23) were tested. The sensitivity and the specificity of the assay were 70.1% and 98%, respectively, when 50 random samples of the 1995 Multiple Autoantibody Workshop were tested (24).
GAD65 ab were measured by a radiobinding assay using 35S-labeled (in vitro transcription/translation) GAD65 as previously described (13). The cut-off index for positivity was 0.025. In the First (25) and Second (Gouvieux Chantilly, France, 1994) International Workshops for the standardization of GAD antibody analysis, the specificities of the method were 100% and 97%, and the sensitivities were 100% and 83%, respectively. Of the sera from the 45 blood donors and 64 healthy Finnish children tested concomitantly with patient sera, 2.2% and 3.1% were positive for GAD65 ab, respectively.
Insulin autoantibodies
IAA were quantified with a microassay (26) modified from that described by Williams et al. (27). Antibody-antigen complexes were precipitated with protein A-Sepharose CL-4B (Pharmacia Biotech) after incubation of serum samples with mono-[125I]TyrA14-human insulin (Amersham Pharmacia Biotech, Little Chalfont, UK) at 4 C for 72 h in the absence or presence of an excess of unlabeled insulin. The IAA levels representing specific binding were expressed in RU based on a standard curve run on each plate using the MultiCalc software program (Perkin-Elmer Corp., Wallac, Inc., Turku, Finland). A subject was considered to be positive for IAA when the specific binding exceeded 1.55 RU (99th percentile in 374 nondiabetic Finnish subjects). The disease sensitivity of our microassay was 35%, and the specificity was 100% based on 140 samples derived from the 1995 Multiple Autoantibody Workshop (24).
Islet cell antibodies
ICA were detected by a standard immunofluorescence assay with 4-µm unfixed cryostat sections of human pancreas and a detection limit of 2.5 Juvenile Diabetes Foundation units, which was considered positive (13). Less than 4% of healthy children or adults are positive for ICA by this method.
HLA genotyping
The HLA class II DQA1, DQB1, and DRB1 genes were typed for 67 patients, including the 11 type I diabetes patients for whom genomic DNA was available. As a general population control, 93 consecutive Finnish cadaver kidney donors were typed with the same techniques. In addition, data from large studies of the Finnish population were available (Ilonen, J., personal communication). The genotyping was performed as previously described (13) using PCR and digestion with allele-specific restriction endonucleases for DQA1, PCR, and chemiluminescent oligonucleotide hybridization for DQB1, and group-specific PCR followed by restriction endonuclease digestion for DRB1. Seven DQA1 alleles, 11 DQB1 alleles, and 15 DRB1 types could be identified.
Statistical analysis
For statistical analysis, the Mann-Whitney U test (levels of C
peptide or FPIR), 
2 statistics with Yates’
correction, or Fisher’s exact test were used when appropriate.
Fisher’s exact test was used for the analysis of HLA allele and
haplotype frequencies.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
By the end of 1998 type 1 diabetes presented in 12 (18%; 7 males
and 5 females) of the 68 patients with APECED included in this study.
Their ages at the diagnosis varied from 4.1–45.3 yr (mean, 24.0
± 12.8 yr; Table 1
). Serum samples taken before the clinical onset of
diabetes were available from 11 patients, samples obtained at the time
of clinical onset were obtained from 9 patients. From 1 patient the
first sample was taken 2 yr after the diagnosis. The sample was
positive for GAD65 ab, but not for IA-2 or ICA. As IAA must be analyzed
from samples taken before the start of exogenous insulin therapy, this
patient was excluded from autoantibody comparisons. Of the 11 diabetics
from whom prediabetic samples were available, 4 (36.3%) were positive
for IA-2 ab, 4 were positive for IAA (36.3%), and 5 (45.5%) were
positive for IA-2 ab and/or IAA. Eight (72.7%) patients had GAD65 ab,
and 6 (54.5%) also had ICA (Table 2).
Two patients were negative for all 4 antibodies (Tables 3 and 4).
The prevalence of IA-2 ab or IAA was significantly higher in patients
with type 1 diabetes (4 of 11; 36.3%) than in those without [2 of 48
(4.0%; P < 0.01) or 0 of 48 (P <
0.001), respectively]. The prevalence of IA-2 ab was lower than that
in the control children with newly diagnosed type 1 diabetes (49 of 73;
67.1%; P = 0.05). For IAA it was about the same
(23, 28). However, we did not have serum samples taken
within 2 yr of diagnosis from 1 of the IAA-negative and 2 of the IA-2
ab-negative patients and thus may have missed the antibody-positive
period.
|
|
|
), suggesting that IA-2 ab and IAA
are less sensitive, but more specific, markers for type 1 diabetes
(Table 5). Of the four IA-2 ab-positive
patients, one had IA-2 ab, GAD65 ab, and ICA 8 yr before the clinical
onset of type 1 diabetes at the age of 45 yr, and the antibodies have
remained positive. For the three other patients, IA-2 ab seroconversion
could be established at 6.0–2.0, 4.2–0.5, and 1.3–0.2 yr before the
diagnosis, respectively. The antibodies disappeared within 4 yr after
the diagnosis. Two of the four IAA-positive patients became positive
about 4 yr and two within 1 yr before diagnosis.
|
suggest that with increasing number of
autoantibodies, the risk of type 1 diabetes also increases. APECED patients who remained nondiabetic
Of the 48 nondiabetic patients with APECED for whom serum samples
were available, none had IAA, 2 (4%) had IA-2 ab, 17 (35%) had GAD65
ab, 11 (23%) had ICA, and 31 (65%) had none of the 3 antibodies
(Table 2). Both of the IA-2 ab-positive patients were also positive for
GAD65 ab, and the other was positive for ICA. One of them has had high
levels of IA-2 ab for 14.3 yr. The other was positive for 8.2 yr until
his death at the age of 37 yr. Age at the time of the first IA-2 ab- or
GAD65 ab-positive sample did not differ between the patients with or
without type 1 diabetes (Table 1).
Data for C peptide levels or FPIR were available for 26 and 28
patients, respectively. No statistically significant differences were
seen in these levels between the patients with or without ß-cell
autoantibodies (Table 1).
HLA II genotypes
All but 1 of the diabetic patients were genotyped for DQA1, DQB1,
and DRB1. In most European populations, including the Finnish, the
typical susceptibility haplotypes for type 1 diabetes are
DR4-DQA1*0301- DQB1*0302, or DR3-DQA1*0501- DQB1*0201 (29, 30). Only 3 of the 11 diabetic patients had the former, and 1
had the latter (Table 4). Susceptibility haplotypes did not affect the
prevalence of GAD65 ab in APECED patients. GAD65 ab were found in 6 of
21 patients with and 19 of 38 without these haplotypes. None of the
IA-2 ab- or IAA-positive APECED patients had the risk haplotypes. No
particular genotype or allele was predominant among the diabetic
patients. Remarkably, however, none of them had DR15 or DQB1*0602
alleles, which are considered protective against type 1 diabetes
(31), whereas 15 of 56 (26.8%) of the nondiabetic
patients (P = 0.047) and 24 of 93 (25.8%) of the
control group (P = 0.047) had these alleles. The
prevalence of DQB1*0602 in all APECED patients (15 of 67; 22.4%) was
similar to that seen in general Finnish population (28%; Ilonen, J.,
personal communication). Among the 59 APECED patients for whom we had
both HLA and autoantibody data, 13 had the DQB1*0602 allele. Four
(30.8%) of them had GAD65 ab, and 13 of 46 patients without DQB1*0602
(30.2%) had these antibodies. Also, no patient with these alleles
tested positive for IA-2 ab, although 4 had both GAD65 ab and ICA.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The destruction of the ß-cells is a chronic process heralded by circulating autoantibodies. IAA, ICA, GAD65 ab, IA-2 ab (ICA512 ab), or phogrin ab have been detected in 80–90% of patients with recent-onset type I diabetes (16, 32). The pathogenetic role of these antibodies is unknown, but they are valuable in the identification of high risk individuals. Patients with APECED are at a very high risk of type 1 diabetes, as 18% of our patients have progressed to clinical diabetes. The few studies on IAA, GAD65 ab, IA-2 ab, or ICA in autoimmune polyendocrine syndromes have been performed mainly in patients with autoimmune polyendocrine syndrome type II (APS II) (33), which usually occurs in adult females and is associated with the same HLA alleles as isolated type 1 diabetes (34). In APS II GAD ab and ICA are not good predictive markers for type 1 diabetes, but IA-2 ab seem to perform better (18, 33, 35). In our previous study of patients with APECED, GAD65 ab were a sensitive marker for type 1 diabetes, and the negative predictive value was as high as 90% (13). However, neither GAD65 ab nor ICA were specific for type 1 diabetes, and their positive predictive value was only 27%. In the present study of 60 patients with APECED, the positive predictive values of IA-2 ab and IAA for type 1 diabetes were 67% and 100%, respectively, whereas the negative predictive values were 85% and 87%. Altogether, we found IA-2 ab and IAA in 36%, GAD65 ab in 73%, and ICA in 55% of the diabetics and in 4%, 0%, 35%, and 23%, respectively, of the nondiabetics. The low frequency of IA-2 ab is not due to an insensitive assay, as 67% of newly diagnosed children with isolated type 1 diabetes were positive by our method. This compares well with the 55–75% ab frequency detected by others in similar patients (28). Only 1% of control children and adolescents were positive in our assay, agreeing well with the less than 2.5% prevalence found by others (28). In fact, the IA-2 ab prevalence of 67% in our ICA-positive patients with type 1 diabetes is in concordance with the frequencies obtained previously (18). All in all, the prevalence patterns of IA-2 ab and IAA in patients with APECED resemble those reported in patients with APS II (18, 33, 35), other organ-specific autoimmune diseases (20), and stiff man syndrome (18, 33). The prevalence of GAD65 ab (43%) in APECED seems to be lower than that in the stiff man syndrome [11 of 14 (79%) (36) or 25 of 28 (89%) (33)].
Both in the first degree relatives of patients with type 1 diabetes and in the general population, the simultaneous presence of two or more of IAA, GAD65 ab, IA-2 ab, and ICA has been reported to be a more specific predictive marker for type 1 diabetes than any of the antibodies alone (16, 37, 38). In patients with APECED, no antibody combination improved the predictive value of IA-2 ab or IAA alone.
The presence of GAD65 ab, IA-2 ab, and ICA in the nondiabetic patients may be a sign of subclinical insulitis. In our previous studies we found an association of GAD65 ab with lower C peptide levels in the nondiabetic patients with APECED (13) or with lower maximal insulin secretory capacity in patients with clinically diagnosed type 2 diabetes (39). In this study we did not find significant differences in the C peptide levels or FPIR values between the GAD65 ab-positive or -negative nondiabetic patients with APECED. This agrees with the recent report by Klemetti et al. (40), who did not find any association between the T cell responses to GAD65 protein and decreased iv glucose tolerance test insulin responses in nondiabetic patients with APECED. It seems that in APECED the IA-2 ab do not always predict the rapid onset of type I diabetes. Two of our 48 nondiabetic patients had very high IA-2 ab levels for more than 8 yr without any functional signs of diabetes. A similar observation has been reported in patients with stiff man syndrome (33).
The HLA genes, especially the class II DQ and DR genes, account for at least 40% of the familial aggregation of IDDM (41). In most Caucasoid populations DR3-DQA1*0501-DQB1*0201, DR4-(DRB1*0401)-DQA1*0301-DQB1*0302, and DR4-(DRB1*0402)-DQA1*0302-DQB1*0302 are high risk haplotypes. As we have also shown (13), the frequency of high risk haplotypes was not increased among APECED patients with type 1 diabetes.
Other HLA DQ alleles are associated with protection from type I diabetes (29, 42). The DQB1*0602 allele occurs in 20–25% of subjects representing the general populations (28% in Finland; Ilonen, J., personal communication), but in less than 1% of children who develop type 1 diabetes (43). The protective effect of DQB1*0602 may even be dominant over susceptibility encoded by the high risk DQB1 alleles (29). Thus, the haplotype DR2 (DRB1*1501)-DQA1*0102-DQB1*0602 appears to protect ICA-positive first degree relatives of patients with type 1 diabetes (31, 44). In the autoimmune syndromes associated with type 1 diabetes the protective effect of the haplotype seems to vary. In stiff man syndrome the DQB1*0602 allele apparently protects the patients (45, 46), but the protection is not absolute (36). In APS II the frequency of DQB1*0602 is not significantly decreased among patients with type 1 diabetes compared with the nondiabetic patients (34). Based on a report of one diabetic patient with APECED it has been suggested that DQB1*0602 would not be protective in APECED (41). In our study none of the tissue-typed diabetic patients with APECED carried the protective DQB1*0602 allele, in contrast to 15 of 56 (26.8%) of the nondiabetic patients (P = 0.047) and 24 of 93 (25.8%) of the control group (P = 0.047). The frequency of this allele in our patients with APECED (22.4%) did not differ from that in the general Finnish population. DQB1*0602 did not protect the patients from GAD65 ab, but none of the DQB1*0602-positive patients had IA-2 or IAA. This agrees with previous findings (44) that the DQB1*0602 allele limits the antibody response mainly to GAD ab.
The results of our study suggest a negative association of the protective DQB1*0602 allele with type 1 diabetes in APECED. In contrast, no positive association with susceptibility haplotypes was seen. This is remarkable, as until now no association of the HLA type II antigens with any disease component of APECED has been reported. The mechanisms by which particular HLA class II alleles mediate susceptibility to autoimmune diseases are unknown (47). In nonobese diabetic mice protection against type I diabetes may in some cases be due to negative selection of autoreactive T cells (48) or positive selection of T cells with suppressive action on the diabetic process (49). Our results suggest that the loss of AIRE gene products does not affect the mechanisms by which HLA class II alleles mediate protection. AIRE messenger ribonucleic acid and protein are expressed in the thymic medulla in cells coexpressing cytokeratin and the dendritic cell markers CD11c and CD83 (7). This part of the thymus is responsible for negative selection of the T cell repertoire (50). It may be that the protective HLA class II alleles, which were reported to bind autoantigenic peptides with high affinity (51), manage to induce tolerance in cytotoxic T cells in either the thymus or the periphery even in the absence of AIRE. Further research is needed to clarify at which level the lack of AIRE affects the induction of autoimmunity or tolerance.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received January 22, 2000.
Revised August 15, 2000.
Accepted September 12, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N. G. Gavalas, E. H. Kemp, K. J. E. Krohn, E. M. Brown, P. F. Watson, and A. P. Weetman The Calcium-Sensing Receptor Is a Target of Autoantibodies in Patients with Autoimmune Polyendocrine Syndrome Type 1 J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2107 - 2114. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Perheentupa Autoimmune Polyendocrinopathy-Candidiasis-Ectodermal Dystrophy J. Clin. Endocrinol. Metab., August 1, 2006; 91(8): 2843 - 2850. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S Ronkainen, T. Harkonen, J. Perheentupa, and M. Knip Characterization of the humoral immune response to glutamic acid decarboxylase in patients with autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) and/or type 1 diabetes Eur. J. Endocrinol., December 1, 2005; 153(6): 901 - 906. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Soderbergh, A. G. Myhre, O. Ekwall, G. Gebre-Medhin, H. Hedstrand, E. Landgren, A. Miettinen, P. Eskelin, M. Halonen, T. Tuomi, et al. Prevalence and Clinical Associations of 10 Defined Autoantibodies in Autoimmune Polyendocrine Syndrome Type I J. Clin. Endocrinol. Metab., February 1, 2004; 89(2): 557 - 562. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Gylling, E. Kaariainen, R. Vaisanen, L. Kerosuo, M.-L. Solin, L. Halme, S. Saari, M. Halonen, O. Kampe, J. Perheentupa, et al. The Hypoparathyroidism of Autoimmune Polyendocrinopathy-Candidiasis-Ectodermal Dystrophy Protective Effect of Male Sex J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4602 - 4608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Anderson, E. S. Venanzi, L. Klein, Z. Chen, S. P. Berzins, S. J. Turley, H. von Boehmer, R. Bronson, A. Dierich, C. Benoist, et al. Projection of an Immunological Self Shadow Within the Thymus by the Aire Protein Science, November 15, 2002; 298(5597): 1395 - 1401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Betterle, C. Dal Pra, F. Mantero, and R. Zanchetta Autoimmune Adrenal Insufficiency and Autoimmune Polyendocrine Syndromes: Autoantibodies, Autoantigens, and Their Applicability in Diagnosis and Disease Prediction Endocr. Rev., June 1, 2002; 23(3): 327 - 364. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Halonen, P. Eskelin, A.-G. Myhre, J. Perheentupa, E. S. Husebye, O. Kampe, F. Rorsman, L. Peltonen, I. Ulmanen, and J. Partanen AIRE Mutations and Human Leukocyte Antigen Genotypes as Determinants of the Autoimmune Polyendocrinopathy-Candidiasis-Ectodermal Dystrophy Phenotype J. Clin. Endocrinol. Metab., June 1, 2002; 87(6): 2568 - 2574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Lethagen, U.-B. Ericsson, B. Hallengren, L. Groop, and T. Tuomi Glutamic Acid Decarboxylase Antibody Positivity Is Associated with an Impaired Insulin Response to Glucose and Arginine in Nondiabetic Patients with Autoimmune Thyroiditis J. Clin. Endocrinol. Metab., March 1, 2002; 87(3): 1177 - 1183. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
