Identification of Novel Predictive Biomarkers for Endometrial Malignancies: N-Acylethanolamines

Objective: To identify new biochemical markers for endometrial cancer (EC). Recent evidence suggests that members of the endocannabinoid system (N-acylethanolamines) that bind to and activate receptors that are dysregulated in EC are involved in this tumour's biology. These observations suggest increased N-acylethanolamine levels in the tissue that might appear in plasma and could be used as disease biomarkers. Methods: N-arachidonoylethanolamine (anandamide, AEA) and the N-acylethanolamine substances, N-oleoylethanolamine (OEA), and N-palmitoylethanolamine (PEA) were quantified in plasma and endometrial tissue collected from 31 EC and seven atrophic controls using UHPLC-MS/MS. Receiver-operating characteristics (ROC) and logistic regression were used to determine diagnostic accuracy. Cannabinoid receptor 1 (CB1) and 2 (CB2) protein levels were determined by specific immunohistochemistry and histomorphometric analyses. Correlations between plasma and tissue levels of the three N-acylethanolamines and tissue levels of the three N-acylethanolamines and CB1 and CB2 receptor expression levels were determined using correlation analysis. Results: Plasma and tissue AEA and PEA levels were significantly (p < 0.05) higher in EC than controls whilst OEA levels were significantly elevated in type 1 EC tissues but not in plasma. There were significant positive correlations between plasma and tissue levels of AEA (R(2) = 0.302, p = 0.008) and PEA (R(2) = 0.182, p = 0.047), but not for OEA (R(2) = 0.022, p = 0.506). The diagnostic accuracies for EC were: sensitivity of 53.3%, specificity of 100% for plasma AEA (>1.36 nM); sensitivity of 73.3%, specificity of 100% for plasma PEA (>27.5 nM); and sensitivity of 93.3%, specificity of 28.6% for plasma OEA (>4.97 nM). Logistic regression increased the area under the ROC curve (AUC) from 0.781 for AEA, 0.857 for PEA, and 0.543 for OEA to a combined AUC of 0.933 for EC diagnosis. Significant inverse correlations between tissue AEA (R(2) = 0.343, p = 0.003) and PEA (R(2) = 0.384, p < 0.0001) levels and CB1 expression were observed. No correlation between tissue levels of OEA and CB1 and tissue levels of any of the three N-acylethanolamines and CB2 protein expression were observed, except in the type 1 EC patients. Conclusion: Since plasma AEA and PEA are significantly elevated in patients with EC and a reflection of production by the endometrial tumour, then these lipids have the potential to be useful biomarkers for the early diagnosis of EC.