Development of molecular diagnostics for nodular thyroid diseases


It is accepted that about 5–10% of the population will develop a clinically significant thyroid nodule during their lifetime. One of the major challenges for the endocrinologists is the evaluation of these thyroid nodules. Although most thyroid nodules are benign, the most common indication for surgical intervention is to exclude the diagnosis of carcinoma.

The current criteria for surgery decisions rely on cytopathologic evaluation of cells received from the fine needle aspiration (FNA), a technique that was introduced as a preoperative test for thyroid nodules in the 1970’s and validated as a reliable test in numerous studies since that time. With the adoption of thyroid FNA, the likelihood of requiring surgery for a thyroid nodule decreased about 50%, from 65–70% to 35–40%, with a concomitant decrease in the number of thyroidectomies performed. Although FNA is currently considered to be the best initial diagnostic test for evaluation of a thyroid nodule, it cannot discriminate between benign and malignant follicular thyroid tumors.  Εven experienced cytologists encounter problems with FNA samples for which cytological features neither confirm nor rule out malignancy. Patients seen with follicular thyroid lesions are advised to undergo surgery of thyroidectomy. In essence, doctors and patients have to decide preoperatively whether to undergo total thyroidectomy based on inadequate clinical information. Especially the cases designated as suspicious or indeterminate lead to thyroid surgery. Current estimates indicate that carcinomas of the thyroid are ultimately found in 10–20% of lesions read as follicular tumors by FNA cytology. As a consequence, diagnosis of cancer is confirmed by pathological examination of surgical pieces in only about one third of patients operated for cancer or suspicion of cancer, which means that the other two thirds of operated patients could have avoided the surgery.

Thus, the current practices for managing a patient with a thyroid nodule deemed a follicular lesion by FNA are confusing and seemingly arbitrary. There is a clear need to develop more accurate initial diagnostic tests for thyroid nodules, particularly for those nodules classified as follicular lesions on initial cytopathological review and potentially to direct subsequent treatment as well.

The first route of this study is to take advantage of all the accumulated information that has been produced from molecular analyses of follicular thyroid nodules by microarray and mutation studies and translate it into clinical use by developing a fast molecular assay that will allow accurate diagnosis and direct both patient’s initial treatment as well as follow-up by a simple sampling with fine needle. There are two types of molecular analysis that we will incorporate in this study. The first is to uncover a standard molecular profile by gene expression analysis of gene-targets described above, that would allow us to discriminate between different pathologies with accuracy and certainty. The second is to determine the mutation status of certain key oncogenes that will supplement the gene expression analysis data and aid in the detailed characterization of a nodule. More specifically, we will look at a panel of mutations that account for most common mutations in differentiated thyroid cancer. It must be emphasized that we will apply high resolution melting (HRM) a new methodology for mutation scanning which is carried out in the same tube or well in which the gene sequence is amplified. It is a new ultra-fast method for mutations detection in a single run of less than 2 hours resulting in extremely rapid screening.

The second route of this proposal is the delineation of new intracellular signaling pathways that control thyroid nodules fate. This part of the study will 1) confirm the results of the gene expression studies at the protein level and 2) uncover potential mechanisms of thyroid pathology. We will obtain surgical specimens of different pathologies and /or we will generate new cell lines from patients with different types of thyroid nodules to study signaling pathways that become altered in thyroid nodules pathologies. We are particularly interested in the role of integrins and their downstream signaling in thyroid function. There is now substantial evidence that certain integrins such as integrin α6β4 (Nikolopoulos et al., 2004; 2005) cooperate with receptor tyrosine kinases to modulate cell proliferation and migration in normal cells and that augmented cooperative signaling by these integrins and associated receptor tyrosine kinases contributes to the pathologic state of the thyroid gland (Ilario et al., 2003). Uncovering relevant pathways and molecular mechanisms contributing to thyroid nodules pathologic type, will result in the identification of new targets of diagnostic, prognostic or therapeutic value and aid in the clinical management of this disease.

Molecular classification of tissues is an emerging technology that will undoubtedly change the way patients are managed in the future. This research plan uses this technology in order to provide a reliable method for the molecular diagnosis of thyroid nodules. The aim of our study is to identify a gene signature that enhances the sensitivity and the specificity of diagnosis of thyroid nodules when compared to traditional cytological techniques and provide a mechanistic rationale by uncovering relevant signalling pathways. Current diagnosis is conventionally obtained by morphological examination of tissue biopsy sections but is not totally reliable due to the nature of the thyroid tissue. Nevertheless, intra- and inter-observer incongruities may occur. The RT-qPCR gene profiling method based on the gene signature described here combined with mutational analysis will be an accurate tool for molecular diagnosis of certain pathologies of thyroid such as the presence/absence of cancerous or pre-cancerous cells not subjected to intra- and inter-observer error. The method would help to enhance diagnosis, particularly in those cases in which it is necessary to resolve uncertainties. In addition, our proposed methodology based on detection of a multiparametric expression signature is expected to yield the best performance at detection compared to single predictors. The method is also anticipated to have potential prognostic value, because the expression level of these genes will correlate with the final outcome of thyroid patients. It is anticipated that information like this obtained from this method will help to tailor treatment based on the patients’ risk profiles.