Thyroid cancer is a common endocrine malignancy with a continuously growing incidence over the last several decades in the United States. However, most of thyroid cancers are indolent and rarely result in patient death. As a result, recent patient management guidelines advocate a more limited surgery (removal of only one thyroid lobe) and overall less aggressive treatment of patients with thyroid cancer that has low to intermediate risk for disease progression.

Importantly, most recent molecular tools can help establish the probability of cancer in thyroid nodules with high accuracy, as well as predict how aggressive the cancer is likely to be. However, the latter requires the analysis of cells collected from thyroid nodules using fine needle aspiration (FNA) for a large group and various classes of genetic mutations that occur in thyroid cancer. This could not be achieved until recently as the genetic tools needed to simultaneously interrogate a large number of genes for multiple alterations types were missing.

The situation changed with the introduction of next generation sequencing. The availability of this technology, paired with increasing experience in detecting various genetic alteration types by early adopters of this technology, makes the prediction of thyroid cancer aggressiveness possible. In Cancer, my colleagues and I reported the results of a study validating the accuracy of an assay called ThyroSeq v3 Genomic Classifier to detect five different classes of molecular alterations: single nucleotide point mutations, insertions/deletion (also called indels), gene fusions, copy number alterations, and abnormal gene expression. Using a large series of thyroid samples representing all major types of thyroid cancer, the authors showed that all types of alterations could be detected preoperatively when the sample collected from a given thyroid nodule contained at least 12% of the tumor’s cells.

Our results suggest that clinicians can accurately predict the probability of thyroid cancer in thyroid nodules, particularly those with indeterminate FNA cytology, as well as identify which cancers have the highest potential to harm the patient. This information, available preoperatively, should help offer a more individualized approach to the surgical and post-surgical treatment of patients with thyroid cancer.


Nikiforova M, Mercurio S, Wald A, et al. Analytical performance of the ThyroSeq v3 genomic classifier for cancer diagnosis in thyroid nodules. Cancer. 2018;;124:1682‐1690. Available at