Endocrine Abstracts

Objective: Thyroid nodules are the most common thyroid disease and preoperative differentiation of benign and malignant lesions is an important diagnostic purpose. In about 1/3 cases FRA and cytological examination shows indeterminate diagnosis. Deregulation of miRNA expression has been described in a variety of tumors, including thyroid cancer. The aim of our study is to evaluate the association between miRNAs expression and risk of thyroid cancer in patients with thyroid nodules.

Methods: In retrospective case-control study we have studied 112 patients with suspected thyroid cancer (FNA Bethesda III-VI). All patients underwent thyroidectomy with histological verification of the diagnosis. The expression of miR144, miR145, miR146, miR155, miR183, miR199, miR221, miR31, miR551, miR375, miR451 and miR7 were determined using quantitative real-time PCR. Logistic regression and receiver operating characteristic (ROC) curve analysis was used to assess the diagnostic efficacy and build mathematical diagnostic models.

Results: Study group included 54 patients with thyroid cancer and control group of 58 patients with benign nodules. In our study up-regulation of miR146, miR221, miR155, miR375, miR31, miR551 and down-regulation of miR7 and miR145 were significantly associated with thyroid cancer. The most diagnostic value had expression of miR145 with an area under the ROC curve (AUC) value of 0,982 (95%CI: 0,953-1,0), miR146 – 0,945 (95%CI: 0,896-0,994), miR375 – 0,913 (95%CI: 0,855-0,971) and miR7 – 0,900 (95%CI: 0,842-0,958). After logistic regression we build mathematical diagnostic models based on the expression of miR145 and miR7. The AUC value for model was 0,99 (0,96-1,00), the sensitivity (Se) was 98.1% (89.9-100), the specificity (Sp) was 96.6% (88.1-99.6).

Conclusion: The expression of miR145 and miR7 has a high diagnostic value in detecting of thyroid cancerKey words: thyroid cancer, molecular genetic testing, miRNA