We confirm the applicability of a number of recently proposed marker genes for the distinction of benign and malignant thyroid tumor and suggest that their diagnostic usefulness is independent of the iodide supply. We propose that the most discriminative marker set identified in our validation study together with marker combinations proposed by other investigators should now be evaluated in multicenter trials.

Objective: We evaluated three markers (insulin-like growth factor II (IGF-II), cyclooxygenase-2 (COX-2) and ets-1) of thyroid growth stimulation and cell transformation together with a thyroidspecific marker (thyroglobulin (Tg)) for their potential to differentiate benign and malignant follicular thyroid neoplasia (FN). Design and methods: mRNA expression levels were determined by real-time PCR in 100 snap-frozen thyroid samples: 36 benign thyroid nodules with different histology and function (19 cold (CTN) and 17 toxic thyroid nodules (TTN)), 36 corresponding normal thyroid tissues of the same patients, eight Graves' disease (GD) thyroids, 10 follicular thyroid carcinomas (FTC) and 10 papillary thyroid carcinomas (PTC). Results: Mean IGF-II and COX-2 levels were not significantly altered between benign and malignant thyroid nodules (IGF-II) or nodular (FTC, TTN, CTN) and normal thyroid tissues (COX-2). In contrast, eight-to tenfold upregulation of ets-1 was observed in PTC and three-to fourfold upregulation of ets-1 was observed in FTC (and GD) compared with benign thyroid nodules and normal thyroid tissues. In addition, thyroglobulin mRNA expression was markedly downregulated (50-to 100-fold) in FTC, PTC and GD samples compared with benign nodular and normal thyroid tissues. Hence an ets-1/Tg ratio . 20 distinguished differentiated thyroid cancer from benign nodular or normal thyroid tissue. We then studied ets1-and Tg mRNA expression levels in fine needle aspiration cytology (FNAC) samples. However, in a consecutive series of 40 FNAC samples only equivocal results were obtained on 38 benign and two malignant (FTC) thyroid tumour samples. Conclusions: Upregulation of ets-1 and downregulation of Tg mRNA expression occur in differentiated thyroid cancer and may facilitate pre-operative identification of thyroid malignancy depending on further evaluation of these potentially promising markers in a larger series of benign and malignant thyroid tumours and their FNAC samples. 152 785-790 European Journal of Endocrinology

Increased oxidative stress has been linked to thyroid carcinogenesis. In this paper, we investigate whether oxidative DNA damage and DNA repair differ in follicular adenoma (FA) and follicular thyroid carcinoma (FTC). 7,8-Dihydro-8-oxoguanine (8-OxoG) formation was analysed by immunohistochemistry in 46 FAs, 52 FTCs and 18 normal thyroid tissues (NTs). mRNA expression of DNA repair genes OGG1, Mut Y homologue (MUTYH) and endonuclease III (NTHL1) was analysed by real-time PCR in 19 FAs, 25 FTCs and 19 NTs. Induction and repair of oxidative DNA damage were studied in rat FRTL-5 cells after u.v. irradiation. Moreover, activation of DNA damage checkpoints (ataxia telangiectasia mutated (ATM) and H2A histone family, member X (H2AFX (H2AFX))) and proliferation index (MIB-1) were quantified in 28 non-oxyphilic and 24 oxyphilic FTCs. Increased nuclear and cytosolic 8-OxoG formation was detected in FTC compared with follicular adenoma, whereby cytosolic 8-OxoG formation was found to reflect RNA oxidation. Significant downregulation of DNA repair enzymes was detected in FTC compared with FA. In vitro experiments mirrored the findings in FTC with oxidative stress-induced DNA checkpoint activation and downregulation of OGG1, MUTYH and NTHL1 in FRTL-5 cells, an effect that, however, was reversible after 24 h. Further analysis of FTC variants showed decreased oxidative DNA damage, sustained checkpoint activation and decreased proliferation in oxyphilic vs nonoxyphilic FTC. Our data suggest a pathophysiological scenario of accumulating unrepaired DNA/RNA damage in FTC vs counterbalanced DNA/RNA damage and repair in FA. Furthermore, this study provides the first evidence for differences in oxidative stress defence in FTC variants with possible implications for therapeutic response and prognostic outcome.

Fine needle aspiration cytology (FNAC) is widely recommended as an important tool for pre-operative identification of malignancy in patients with nodular thyroid disease. To assess the diagnostic contribution of FNAC and the potential of quantitative mRNA analysis in fine needle aspirates in daily practice, we conducted a prospective study in thyroid clinics (n=2) and endocrine practices (n=3), respectively in an East German region with borderline iodine deficiency. Two-hundred and forty-four consecutive FNACs were obtained over a period of 2 years (2002-2004) from euthyroid patients presenting for first evaluation of a solitary thyroid nodule. The mean nodule size for FNAC was 27 mm (range: 10-79 mm). In 55% of patients FNAC was performed after scintiscan detection of a cold or normal functioning thyroid nodule (CTN), while in the remainder FNAC was performed as a primary investigation. FNAC outcomes were: 57.8% benign, 22.1% indeterminate, 2.5% suspicious for malignancy, 17.6% non-diagnostic. Messenger RNA levels for a house keeping gene (beta-actin) and a thyroid specific marker (thyroglobulin, Tg) were studied as basic molecular markers using real-time PCR. Both in the IN VIVO and EX VIVO FNA series, beta-actin and Tg mRNA levels were positively correlated with the thyrocyte cell yield/respective FNA smear. However, subgroup analysis showed that FNAC with histologically confirmed follicular thyroid cancer and/or microfollicular adenoma exhibited significantly lower Tg mRNA expression despite high beta-actin levels. Sufficient mRNA quantities were obtained in >90% of FNA specimen to allow quantitative mRNA analysis of at least 5 further genes. In conclusion, quantitative mRNA analysis is feasible in FNA on a routine basis and provides a perspective for a molecular distinction of thyroid nodules, once specific marker genes have been defined for benign and malignant thyroid tumours respectively.

The levels of genomic DNA were significantly decreased in GAA treated vs non-treated TP samples. Aliquoting from the TP sample prior to treatment with GAA enables accurate measurement of DNA without affecting the adequacy of the TP cytology slide.