Date of Award

3-22-2021

Document Type

Doctoral Dissertation - Restricted (NYMC/Touro only) Access

Degree Name

Doctor of Philosophy

Department

Basic Medical Sciences

First Advisor

Jan Geliebter

Abstract

Risk stratification is a cornerstone of the clinical management of papillary thyroid cancer (PTC). Molecular markers that quickly and accurately predict aggressive features of complicated or refractory disease may inform management at an early stage. Currently, most molecular markers in PTC are based on genomic or transcriptomic alterations of protein-coding genes and microRNAs. Furthermore, current knowledge of PTC oncogenesis does not explain the higher incidence of malignancy in women relative to that of men. Previously, we hypothesized that variation in sex hormone signaling may underlie sex-bias in follicular cell thyroid cancers. Specifically, we postulated that the androgen receptor (AR), a potent transcription factor at several regulatory DNA elements with distinctly lower activity in females, drives the differences in the PTC immune landscape between males and females. In this study we present a large, novel PTC transcriptomic and epigenome analysis to investigate how transcriptional regulators, including AR and long-noncoding RNA (lncRNA), are associated with clinicopathological characteristics.

We performed whole-transcriptome RNA-sequencing on 45 primary PTC tumors with matched, normal tissues. Using thyroid differentiation score, weighted gene co-expression network analysis and hallmark gene set enrichment analysis, we identified lncRNAs that are highly predictive of molecular differentiation status and lymph node metastasis.

We identified the FAM95C lncRNA as correlated with thyroid differentiation (R2 = 0.64), downregulated in BRAFV600E tumors and uniquely expressed among endocrine tissues. In a cluster of co-expressed genes associated with epithelial-mesenchymal transition and TNFα signaling, we isolated 74 lncRNAs highly predictive of lymph node metastasis (p = .05). Finally, we performed Kallisto transcript quantification using a custom index to maximize lncRNA detection and provide a user-friendly, interactive method for data exploration.

Our results provide evidence that lncRNAs may serve as promising biomarkers for predicting clinically aggressive features that guide clinical management such as poor differentiation and lymph node metastasis. To promote the development of novel, more robust molecular markers in PTC, we provide an easy-to-use interface for investigators to access our transcriptomic dataset.

Share

COinS