ABSTRACT: Although fine needle aspiration (FNA) cytology is helpful in determining whether thyroid nodules are benign or malignant, this distinction remains challenging in follicular neoplasms. Identification of genomic alterations in cytological specimens that distinguish between benign and malignant follicular tumors would have great clinical value. This study identified chromosomal imbalances in a set of 126 thyroid frozen follicular lesions characterized by histopathology, using aCGH, FISH, and DNA ploidy determination. An evaluation algorithm was developed based on a panel of three FISH experiments to better distinguish between follicular adenomas and carcinomas of the thyroid. This algorithm detects structural abnormalities that were observed in about half of the follicular thyroid carcinomas. These molecular markers may be therefore useful as an adjunct for making a differential diagnosis of follicular adenoma versus follicular carcinoma using cytological material obtained by FNA. Further information on the values in the comment columns: [FISH_1p_Loss : Status for abnormality found in terms of copy number variation for chr1p by FISH], [FISH_22_Loss : Status for abnormality found in terms of copy number variation for chr22 by FISH], [FISH_X_Gain : Status for abnormality found in terms of copy number variation for chrX by FISH], [FISH_Global : Status for abnormality found in terms of copy number variation for at least one of these (chr1p, chr22, chrX) by FISH], [DNA_Index : Measured by flow cytometry], [KRAS_Mutation : Mutation status for KRAS, obtained from Sanger sequencing. wt for wild-type, mut for mutant], [HRAS_Mutation : Mutation status for HRAS, obtained from Sanger sequencing. wt for wild-type, mut for mutant], [NRAS_Mutation : Mutation status for NRAS, obtained from Sanger sequencing. wt for wild-type, mut for mutant], [BRAF_Mutation : Mutation status for BRAF, obtained from Sanger sequencing. wt for wild-type, mut for mutant], [CTNNB1_Mutation : Mutation status for CTNNB1, obtained from Sanger sequencing. wt for wild-type, mut for mutant], [PIK3CA_Mutation : Mutation status for PIK3CA, obtained from Sanger sequencing. wt for wild-type, mut for mutant], [PAX8-PPARg_Fusion : Status for the PAX8-PPARG gene fusion, obtained from qRT-PCR. pos for presence of fusion, neg for absence], [RET-PTC_Fusion : Status for the RET-PTC gene fusion, obtained from qRT-PCR. pos for presence of fusion, neg for absence], [Mutation_Status : Global mutation status, combining results from HRAS, NRAS mutation status, and PAX8-PPARG fusion status], [Tumor_Size_cm : Tumor size in centimeters], [pT : Classes defined according to the size of the tumors (for more information, check AJCC classification)], [pN : Classes defined according to the number of nodes involved or not (for more information, check AJCC classification)], [pM : Classes defined according to the presence or absence of distant metastases (for more information, check AJCC classification)], [Carcinoma_Histo_Type : Histological type for carcinoma. MIF for minimally invasive, WIF for widely invasive], [Adenoma_Histo_Type : Histological type for adenoma. MIC for microvesicular, MAC for macrovesicular, BOTH for both], [Follow-up_month : Patient follow-up in months], [Alive_status : Patient alive status. yes for alive patient, no for deceased patient], [Metastasis_status : Presence of metastasis during the follow-up. yes for positive, no for negative, possible when uncertainty remains].