Clinical Endocrinology (2008) 69, 664–668 doi: 10.1111/j.1365-2265.2008.03245.x © 2008 The Authors 664 Journal compilation © 2008 Blackwell Publishing Ltd ORIGINAL ARTICLE Blackwell Publishing Ltd Epitope recognition patterns of thyroid peroxidase autoantibodies in healthy individuals and patients with Hashimoto’s thyroiditis* Claus H. Nielsen*, Thomas H. Brix†, Andrzej Gardas‡, J. Paul Banga§ and Laszlo Hegedüs† *Department of Clinical Immunology, Section 7631, Rigshospitalet University Hospital, Copenhagen, Denmark, †Department of Endocrinology and Metabolism, Odense University Hospital, Odense, Denmark, ‡Department of Biochemistry, Medical Centre of Postgraduate Education, Warsaw, Poland and §Division of Gene and Cell-based Therapy, King’s College London School of Medicine, London, UK Summary Objective Thyroid peroxidase antibodies (TPOAb) are markers of autoimmune thyroid disease (AITD), including Hashimoto’s thyroiditis (HT), but naturally occurring TPOAb are also detectable in healthy, euthyroid individuals. In AITD, circulating TPOAb react mainly with two immunodominant regions (IDR), IDR-A and IDR-B. The present study was undertaken in order to compare the epitope recognition pattern of TPOAb in HT patients and healthy subjects. Design Sera from 21 out of 98 healthy controls were selected on the basis of high TPOAb values, required for determination of TPOAb recognition pattern; as were sera from 92 HT patients. Measurements Measurement of IDR-reactivity was possible in 90 patients and 12 controls. IDR-A-, IDR-B- and non-IDR-A/non-IDR- B-Ab constituted 24 ± 11%, 50 ± 15% and 26 ± 12%, respectively, in the patients. The distribution in the controls was distinctly different, only 12 ± 13% being directed against IDR-A (P < 0·002) and 66 ± 22% against IDR-B (P < 0·002). Half of the healthy individuals, vs. none of the HT patients, lacked IDR-A reactivity completely (P < 0·0001). In HT patients, IDR-B-Ab proportions increased slightly with increasing TPOAb levels (P < 0·05), while IDR-B-Ab of the controls showed a strong opposite trend (P < 0·0001). Accordingly, the proportion of non-A/non-B-Ab correlated with TPOAb levels in the healthy controls (P < 0·008), and an inverse correlation was seen in HT patients (P < 0·02). Conclusion The data suggest that TPOAb do not differ only in quantity between HT patients and healthy individuals, but may also follow distinct qualitative patterns. Larger studies are required to confirm this, and to determine whether the propensity to produce antibodies to certain TPO epitopes, for example, IDR-A, is of pathogenic relevance. (Received 6 January 2008; returned for revision 21 January 2008; finally revised 28 January 2008; accepted 1 February 2008) Introduction Hashimoto’s thyroiditis (HT) and Graves’ disease (GD), together known as autoimmune thyroid disease (AITD), are classical examples of organ-specific autoimmune conditions. Both GD and HT are almost invariably accompanied by elevated levels of thyroid peroxidase antibodies (TPOAb), and usually by thyroglobulin antibodies (TgAb). 1 It has been a matter of controversy whether these antibodies play an important role in the pathogenesis, or occur merely as an epiphenomenon to thyroid tissue destruction. 2 Recently, it was shown that TPOAb mediate antibody-dependent cytotoxicity as well as complement-dependent cytotoxicity to thyroid cells in culture. 3 It is not known whether certain antibody specificities are particularly harmful in this respect. However, the epitope specificity of soluble antibodies, as well as B cell antigen receptors, can profoundly effect antigen processing by antigen-presenting cells (APCs) and thereby determine which peptides are presented to T cells. 4–6 It is not known whether the TPOAb repertoire of AITD patients differ from that of healthy individuals or patients with non-AITDs, but a recent investigation shows that TgAb from AITD patients exhibited a different recognition pattern from that of patients, with nontoxic multinodular goiter and papillary thyroid carcinoma. 7 The prevalence of clinically overt AITD is around 2% of women and 0·2% of men, but a 10-fold higher number of persons are considered to have subclinical disease, defined as euthyroidism and thyroid autoantibodies in the circulation. 8,9 The cut-off between antibody-positivity and antibody-negativity is usually set arbitrarily at 30–100 IU/ml for both TPOAb and TgAb. The positive/negative dichotomy is somewhat misleading, however, as measurable levels of TPOAb as well as TgAb can be found in most individuals. 10–12 These autoantibodies presumably belong to a group of autoantibodies *See Commentary on page 526. Correspondence: Claus H. Nielsen, Department of Clinical Immunology, The Tissue type Laboratory, Section 7631, Blegdamsvej 9, Rigshospitalet University Hospital, DK-2200 Copenhagen, Denmark. Tel.: +45 35 45 76 31; Fax: +45 35 39 87 66; E-mail: c.h.nielsen@rh.regionh.dk