Large-Scale Chromosomal Derangements in Hurthle Lesions Revealed By Multiple Genome-Wide High Density Platforms

Program: Abstracts - Orals, Poster Previews, and Posters
Session: SAT 270-310-Thyroid Neoplasia (posters)
Saturday, April 2, 2016: 1:15 PM-3:15 PM
Exhibit/Poster Hall (BCEC)

Poster Board SAT 285
Giulia C Kennedy*1, Su Yeon Kim1, Kevin Travers1, Bamboo Lin1, Grazyna Fedorowicz1, Ed Tom1, Mei Wong1, Dan Pankratz1, P. Sean Walsh1, Jing Huang1 and Virginia Anne LiVolsi2
1Veracyte, Inc., South San Francisco, CA, 2University of Pennsylvania School of Medicine, Philadelphia, PA
RATIONALE.Hurthle cell tumors are a class of thyroid neoplasms characterized by cells with oxyphilic (eosinophilic) voluminous cytoplasm due to numerous enlarged and abnormal mitochondria. They may be adenomas or carcinomas; these are differentiated by presence or absence of capsular or vascular invasion.  The goal of this study was to explore the genomic landscape of Hurthle and other thyroid malignancies and to determine if genome-wide data could be used to uncover copy number aberrations and to correlate these findings with histopathology features.

METHODS.We analyzed 55 thyroid tissues across a spectrum of malignant histopathology diagnoses including Hurthle, papillary, medullary, follicular and anaplastic thyroid cancers as well as a variety of benign subtypes such as Hurthle and follicular adenomas, lymphocytic thyroiditis, and nodular hyperplasia. We isolated RNA and DNA and subjected each sample to deep RNA sequencing or DNA Exome sequencing as well as to CytoScan microarray analysis. We used a variety of algorithmic approaches to identify regions of chromosomal gain and loss, with or without mosaicism, copy neutral loss-of-heterozygosity (LOH), and to identify rearrangements and gene fusions.


RESULTS. Chromosome-level aberrations were not observed for most samples, however among 12 samples that show histopathological patterns consistent with Hurthle lesions, 7 (5 carcinomas and 2 adenomas) exhibited striking aberrations affecting more than 15% of their respective genomes. In contrast, among the 43 samples lacking Hurthle features, only 2 (both follicular carcinomas) show such level of aberrations. Among Hurthle lesions, a substantial proportion of the genome is affected, with diverse types of genomic aberrations, including copy number gain (n=5) and loss (n=2); LOH with and without copy number change (n=4), as well as mosaicism (gain: n=2; loss: n=3). Some samples exhibited multiple types of aberrations. A proportion of the changes observed in DNA were also manifest in the RNA sequencing data, suggesting a biological role for these widespread derangements.

CONCLUSIONS. Genome-wide analysis across multiple platforms reveals consistent evidence of genomic instability in a subset of Hurthle lesions. These aberrations do not correlate solely with malignant histopathology but may be linked to features not observable in histopathological review. This information may shed light on the diagnosis of these challenging thyroid neoplasms.

Disclosure: GCK: Chief Scientific Officer, Veracyte, Inc.. SYK: Researcher, Veracyte, Inc.. KT: Researcher, Veracyte, Inc.. BL: Researcher, Veracyte, Inc.. GF: Researcher, Veracyte, Inc.. ET: Researcher, Veracyte, Inc.. MW: Researcher, Veracyte, Inc.. DP: Researcher, Veracyte, Inc.. PSW: Researcher, Veracyte, Inc.. JH: Researcher, Veracyte, Inc.. VAL: Collaborator, Veracyte, Inc..

*Please take note of The Endocrine Society's News Embargo Policy at