Comparison of gene expression profiles of hot and cold thyroid nodules and papillary thyroid carcinoma

Abstract

Up to now microarray technology was used to uncover the gene expression profiles of both, benign and malignant thyroid pathologies, in 17 different studies using different microarray platforms and -generations. A comparison of the data accumulated in these studies across the different platforms and -generations could improve and focus the analysis of these data. Therefore, the aim of this study was to evaluate the comparability of data sets derived from different platform generations, to examine the influence of intra- and inter-individual comparisons on gene expression profiling results, and to reanalyze the data sets of autonomously functioning thyroid nodules (AFTNs), and cold thyroid nodules (CTNs) with previously not applied support vector machines and papillary thyroid carcinoma (PTC) with previously not applied GenMAPP analysis for the expression patterns of biologically relevant gene sets. The comparison of gene expression data generated with different GeneChip generations, revealed a high reproducibility, when the results of independent analyses within a GeneChip generation were compared. Moreover, our results point out another critical point of the analysis of gene expression data: a precise selection of the reference samples is mandatory to identify subtle changes in the expression patterns of signaling cascades, as we could show in case of the MAPK cascade in the PTCs. While differences in the expression pattern of MAPK cascade associated genes were not detectable in a data set comprising 7 PTC and 11 benign thyroid tissue samples, an increased expression of various MAPK cascade associated genes (e.g. H-Ras, R-Ras, MEK kinase 1, MAP kinase kinase 1, and MAP kinase 1) was detectable in two independent data sets comprising 8 PTCs and their normal surrounding tissue and 16 PTC and their normal ST, respectively. Moreover, a molecular PTC classifier built on GeneChips U133A can be successfully applied to samples analyzed by GeneChips U95A: all AFTNs and CTNs were classified as benign/normal thyroid in contrast to additional PTC samples.