Diagnosing the risk of malignancy in thyroid nodules


Title: Microarray analysis of thyroid nodule fine-needle aspirates accurately classifies benign and malignant lesions.

Authors: Lubitz CC, Ugras SK, Kazam JJ, Zhu B, Scognamiglio T, Chen Y-T, Fahey TJ III.

Reference: Journal of Moleculer Diagnostics 8: 490-498, 2006



The frequency of thyroid nodules- detection has, thanks to ultrasound (US) examinations, increased very much since twenty years. However, only a minor and decreasing fraction (a few percent) of these nodules is malignant. To avoid a huge number of unnecessary surgical interventions, full proof diagnosis would be essential. The microarray technology for defining mRNA expression might contribute to this aim as it does for the prediction of severity in the case of breast cancer.


The objective of the study was to assess the validity of microarray analysis of fine needle aspirates (FNA) for the diagnosis of malignancy in thyroid nodules.

Material and methods

The microarray mRNA expression profile of fifty benign thyroid lesions and papillary carcinomas were obtained and compared to the pathological diagnosis (follicular adenomas, hyperplastic nodules, classical papillary carcinomas and follicular variants of papillary carcinomas). From this analysis, a list of twenty five genes with differential expression between benign and malignant lesions (all with -P- value lower than 1%) was drawn. This list, with seven under-expressions and eighteen over-expressions of genes in the carcinomas largely overlaps lists of the literature.


The 25 genes list (training set) was used to compare FNA specimens obtained from in vivo samples immediately after surgery and matched tissue samples. In 7 cases the closest match to the FNA was the corresponding tissue. In 2 cases this was not true, but the classification was correct. In the 22 FNA samples tested, the microarray and the pathological examination of the FNA as benign versus malignant or undeterminate (in 5 cases) coincided.


The results show that the microarray analysis of samples obtained from FNA is technically feasible and might provide an objective result with an easy automated method.


Fine needle aspiration (FNA) is the recognized tool used to decide whether to operate a thyroid nodule or not. Gene expression microarray analysis is becoming the best tool to define molecular signatures of distinct pathologies or for prognosis. It also suggests possible new therapeutic targets and molecular information to understand the physiopathologic processes. Present article shows that microarray analysis of gene expression on amplified mRNA from FNA samples provides an objective – easy to automate – alternative to the more subjective cytological examination. The power of the analysis is the same, as 5 of 22 specimens were noted -undeterminate- by both methods. The main drawback (i.e. the representativity of analyzed samples) is also the same, due to heterogeneity and ambiguity of the lesions.

Two inherent weaknesses in present study were the fact that: a) the study only compared follicular adenomas with papillary carcinomas, a distinction which pathologists find relatively easy and not the much more difficult differential diagnosis between follicular adenomas and carcinomas. This was probably due to the scarcity of follicular carcinomas in the USA; b) we do not know if the hyperplastic nodules were -hot nodules- (i.e. hyperfunctional).

Nevertheless, this is a first study that will probably be extended to more extensive series and allow, as it was the case for breast tumors, to define standardized criteria for diagnosis, prognosis, and treatment choice. We hope that, as is usually the case with this group, the data will be available on the Internet for further independent analysis.

( Summary and commentary prepared by Jacques Dumont )

Present summary and commentary are related to Chapter N- 18 of TDM (Section on FNA)