Context: Despite extensive searches for novel noninvasive diagnostics, laparoscopy remains the reference test for endometriosis. Circulating miRNAs are purported endometriosis biomarkers; however, the miRNA species and their diagnostic accuracy differ between studies and have not been validated in independent cohorts.
Objective: Identify endometriosis-specific plasma miRNAs and determine their diagnostic test accuracy.
Setting: Two university-based, public hospitals and a private gynecology practice in Australia.
Design and participants: Four phases: (i) Explorative phase. Plasma miRNA menstrual cycle fluctuations were evaluated in women with endometriosis and asymptomatic controls (n = 16). (ii) Biomarker discovery. Endometriosis-specific plasma miRNAs were identified in (a) women with endometriosis and asymptomatic controls (n = 16) and (b) women with and without surgically defined endometriosis (n = 20). (iii) Biomarker selection. Plasma miRNAs with the best diagnostic potential for endometriosis were selected in a surgically defined selection cohort (n = 78). (iv) Biomarker validation. The diagnostic test accuracy of these miRNAs was calculated in an independent, surgically defined validation cohort (n = 119).
Results: Forty-nine miRNAs were differentially expressed in women with endometriosis. Nine maintained dysregulation in the selection cohort, but only three (miR-155, miR574-3p and miR139-3p) did so in the validation cohort. Combined, these three miRNAs demonstrated a sensitivity and specificity of 83% and 51%, respectively.
Conclusion: Plasma miRNAs demonstrated modest sensitivity and specificity as diagnostic tests or triage tools for endometriosis. Other groups' findings were not replicated and accorded poorly with our results. Circulating miRNAs demonstrate diagnostic potential, but stringent, standardized methodological approaches are required for the development of a clinically applicable tool.
Copyright © 2019 Endocrine Society.