Mesenchymal tumours of the corpus uteri comprise common benign lesions - leiomyomas and very rare malignant variants - sarcomas. It can be difficult to distinguish between the particular types of mesenchymal tumours pre-surgically. Primarily, leiomyomas and the very aggressive leiomyosarcomas can be easily misdiagnosed when using only imaging devices. Therefore, a reliable non-invasive marker for these tumour types would provide greater certitude for patients that the lesion remains benign. Our collection comprises 76 native leiomyomas, an equal number of healthy myometrium samples and 49 FFPE samples of various types of sarcomas. The methylation level was assessed by MS-HRM method and we observed differences in the methylation level between healthy, benign and (semi)malignant tissues in the KLF4 and DLEC1 genes. The mean methylation levels of leiomyomas compared to myometrium and leiomyosarcomas were 70.7% vs. 6.5% vs. 39.6 % (KLF4) and 66.1% vs. 14.08% vs. 37.5% (DLEC1). The ATF3 gene was differentially methylated in leiomyomatous and myometrial tissues with 98.1% compared to 76.6%. The AUC values of the predictive logistic regression model for discrimination between leiomyomas and leiomyosarcomas based on methylation levels were 0.7829 (KLF4) and 0.7719 (DLEC1). Finally, our results suggest that there should be distinct models for the methylation events in benign leiomyomas and sarcomas, and that the KLF4 and DLEC1 genes can be considered potential methylation biomarkers for uterine leiomyomas.
Keywords: Leiomyomas; Leiomyosarcomas; MS-HRM; Mesenchymal tumours; Sarcomas; Uterus.
Copyright © 2019 Elsevier GmbH. All rights reserved.