Targeted lipidomics unveils complex lipids in cervicovaginal environment as diagnostic and prognostic biomarkers for endometrial cancer

Global incidence of endometrial cancer (EC) is on the rise. EC-associated risk factors like obesity, metabolic syndrome, and genetic predisposition increase a person’s risk for developing EC in their lifetime and negatively affect prognostic outcomes. Present methods for EC diagnosis rely on invasive endometrial biopsies or dilation and curettage. Our previous global untargeted metabolomics analysis of minimally invasive cervicovaginal lavage (CVL) samples revealed an enrichment in simple lipid signatures associated with EC presence and pathophysiology. This study investigated complex lipids beyond those reported previously, with intent to identify novel lipid biomarkers for accurate, minimally invasive EC detection and prognosis. We analysed 223 CVL samples collected by a physician from women diagnosed with EC ( n = 66), benign pathological conditions ( n = 108) and asymptomatic, healthy women ( n = 49). The concentrations (nmol/mL) of 914 individual lipid species, 14 lipid classes and 28 fatty acid (FA) side chains were detected across CVL samples using the Complex Lipid Targeted Panel (Metabolon Inc.). Tumour size, grade, myometrial invasion, and mismatch repair (MMR) status in women with EC were determined from histopathologic reports. MetaboAnalyst and Prism were used for statistical and biomarker analysis which included Receiver Operating Characteristic (ROC) curves, measuring the Area Under the Curve (AUC) and Youden Index (J). Women with EC exhibited significant enrichment of neutral lipid species, especially those containing long chained polyunsaturated fatty acid side chains. Cholesteryl ester (CE) species, specifically CE(22:4), were consistently identified as top predictive features when applied individually and when incorporated into multivariable diagnostic models. A 10-feature model developed in the exploratory dataset, discriminated EC from a combined group of benign and healthy controls with ≥75% sensitivity and specificity when validated using previously internal held-out samples. Top predictive CE species for EC also held prognostic value as they related to MMR status and histologic type. These findings revealed EC-associated alterations in lipid metabolism within the cervicovaginal environment, highlighting the potential use of complex lipid species detected in CVLs as minimally invasive diagnostic biomarkers for EC and markers associated with EC prognostic features.