Methylation markers differentiate thyroid cancer from benign nodules.

Abstract

The incidence of thyroid cancer (TC) is increasing. Cytology by itself cannot distinguish TC from some benign nodules especially in certain subtypes of TC. Our immediate goal is to identify DNA methylation markers for early detection of TC and to molecularly differentiate TC subtypes from benign nodules. Promoter methylation status of 21 candidate genes was examined on formalin-fixed paraffin-embedded tissue (FFPE) utilizing quantitative methylation-specific polymerase chain reaction (QMSP) in a retrospective cohort of 329 patients (56% white, 29% African American, 61% female) comprising 71 normal thyroid, 83 benign nodules [follicular adenomas (FA)], 90 follicular TC (FTC) and 85 papillary TC (PTC). All genes were analyzed individually (Kruskal-Wallis and Wilcoxon rank sum tests) and in combination (logistic regression models) to identify genes whose methylation levels might best separate groups. Combination gene panels TPO and UCHL1 (ROC=0.607, sensitivity 78%) discriminated FTC from FA, and RASSF1 and TPO (ROC=0.881, sensitivity 78%) discriminated FTC from normal. Methylation of TSHR distinguished PTC from FTC (ROC=0.701, sensitivity 84%) and PTC from FA (ROC=0.685, sensitivity 70%). The six gene panel of TIMP3, RARB2, SERPINB5, RASSF1, TPO and TSHR, which differentiates PTC from normal thyroid, had the best combination sensitivity (91%) and specificity (81%) of the panels addressing discrimination of cancer tissue. Aberrant gene methylation used in combination panels may be useful clinically in differentiating FTC and PTC from benign nodules. If confirmed in additional studies, these findings could help reduce the over diagnosis of thyroid cancer and surgeries related to over diagnosis.