MicroRNA profiling reveals potential biomarkers for the early transformation of endometriosis towards endometriosis-correlated ovarian cancer

Abstract

Background: Endometriosis (EMS) is a chronic, gynecological condition affecting 6-10 % of reproductive-age women. While these lesions are benign, ovarian EMS presents cancer-like features, and can progress to endometriosis-correlated ovarian cancer (ECOC) through a multistep process. Given the regulatory role of miRNAs in gene expression and biological pathways, we aimed to identify miRNAs associated with the malignant transformation of ovarian EMS, which could serve as a potential diagnostic tool for the early identification of such patients.

Methods: Global miRNA profiling was performed in 8 patients with benign ovarian EMS (EMS-b) and 29 patients with ECOC. Differential expression analysis (DEA) of miRNAs between EMS-b, EMS tissues from patients with ECOC (EMS-k) and ECOC tissues was performed. Receiver Operating Characteristic (ROC) curves were built to evaluate the binary classification performance of significant miRNAs.

Results: Comparison between EMS-b and EMS-k revealed 13 significantly deregulated miRNAs. Furthermore, when comparing ECOC and EMS-b, we observed significant deregulation of 181 miRNAs. ROC analysis revealed a panel of seven upregulated miRNAs with accuracies above 0.7 in identifying EMS-k and EMS-b. Notably, four miRNAs (hsa-miR-200a-3p, hsa-miR-141-3p, hsa-miR-183-5p, hsa-miR-10a-5p) were consistently upregulated in both EMS-k and ECOC tissues, achieving accuracies above 0.77 in distinguishing between EMS-k and EMS-b. When used to distinguish between EMS-b and ECOC tissues, these miRNAs showed accuracies even higher, above 0.94. Specifically, hsa-miR-183-5p had an accuracy of 1, hsa-miR-200a-3p and hsa-miR-141-3p of 0.97, while hsa-miR-10a-5p of 0.95.

Conclusions: Our study identified a panel of miRNA biomarkers that may serve as potential candidates for the early detection of ECOC in patients previously diagnosed with ovarian EMS.

Keywords: Endometriosis; Endometriosis-correlated ovarian cancer; Predictive biomarkers; miRNAs.

Fig. 1

Study design and workflow. n: number of cases/samples; EMS-b: benign endometriosis; EMS-k: endometriosis from patients with ECOC; ECOC: endometriosis-correlated ovarian cancer.

Fig. 2

A. Volcano plot showing the significantly deregulated miRNAs between EMS lesions from patients with ECOC (EMS-k) and benign EMS (EMS-b). B. ROC curves showing the accuracy of these miRNAs in discriminating between benign EMS (EMS-b) and EMS from ECOC patients (EMS-k). C. Expression levels of miRNAs significantly upregulated in EMS from ECOC patients. EMS-b: benign endometriosis; EMS-k: endometriosis tissue from patients with ECOC; ECOC: endometriosis-corelated ovarian cancer.

Fig. 3

A. Volcano plot showing the significantly deregulated miRNAs between ECOC and benign EMS. B. Expression levels of miRNAs significantly upregulated ECOC patients. C. ROC curves showing the accuracy of these miRNAs in discriminating between benign EMS patients and ECOC patients compared to the performance of currently used markers (CA-125 and CA-19.9). EMS: endometriosis; ECOC: endometriosis-correlated ovarian cancer.

Fig. 4

ROC parameters for the diagnostic power of the miRNA signature with hsa-miR-200a-3p, hsa-miR-141–3p, hsa-miR-183–5p, hsa-miR-10a-5p); th: best discriminatory threshold; EMS-b: benign endometriosis; EMS-k: endometriosis from ECOC patients; Sp: specificity; Se: sensitivity; NPV: negative predictive value; PPV: positive predictive value; ECOC: endometriosis-correlated ovarian cancer.