Meta-Analysis of the Molecular-Expression Signatures Accuracy in Cytological Indeterminate Thyroid Nodules
Julio A Diaz-Perez, Abhinay Reddy, Deirdre Amaro. University of California, San Diego, La Jolla, CA; University of Santander, Bucaramanga, Santander, Colombia
Background: Fine-needle aspiration cytology (FNA) has reliable accuracy with excellent clinical performance in the preoperative diagnosis of thyroid nodules. However, approximately 25% of nodules are classified as indeterminate. Multiple strategies have been developed to address this problem. Recently, molecular based tests have emerged as promising tools for improving diagnostic accuracy in indeterminate cases.
Design: A systematic review of the literature with meta-analysis of studies that evaluate molecular preoperative test for detection of thyroid cancer in FNA indeterminate nodules was performed. The studies search was executed in the database Medline/PubMed. Search terms used include: “cytology,” “indeterminate,” “molecular,” “gene,” “thyroid,” “accuracy,” “performance,” and “cancer”. Forty-two citations were found and reviewed, and only the original validation studies (OVS) were selected.
Results: We identified 10 OVS published between 2006 and September 2012, which compared different molecular preoperative thyroid test results in indeterminate cytology using histopathology results as final diagnosis. All the studies were analyzed as transversal or naturalistic validations. The total population evaluated was 1354 patients. Five studies evaluated gene-expression classifiers, 3 studies evaluated microRNA classifiers and 2 evaluated immunocytochemistry expression. The sensitivity was reported from 37% (BRAFV600E mutation study) to 100% (miRNA and immunocytochemistry studies) with a proportion of variability (R2) of 1.47E-03. The specificity was reported from 20% (expression levels of four miRNAs miR-7, -126, -374a, and let-7g) to 100% (panel of mutations in BRAFV600E, NRAS, HRAS, KRAS RET/PTC1, RET/PTC3, and PAX8/PPAR-gamma) with a R2 of 0.18. A critical review revealed that 6 of the 10 studies had the potential for verification/selection bias and only 5 stated that the molecular test and histopathology were independently assessed.
Conclusions: New molecular strategies for the diagnosis of indeterminate thyroid nodules have been evaluated with variable sensibility and more consistent specificity. These new studies propose different molecular markers with better performances when used as panels. Limitations in the verification of data were found.
Wednesday, March 6, 2013 1:00 PM
Poster Session VI # 44, Wednesday Afternoon