Background Endometrial cancer (EC) is the most common malignancy of the female reproductive tract. Despite years of research, the accurate screening strategy is still not available in this disease and it is usually diagnosed only after the clinical signs are present. The recent technological advances in analytical methodologies enabled detection of multiple molecules in one, small sample of biological materials. Such approach was undertaken in the presented study. Methods Concentrations of aldehyde dehydrogenase 1 family, member A1 (ALDH1A1), carbonic anhydrase IX (CA9), CD44, epithelial cell adhesion molecule (EpCAM), hepsin, kallikrein-6, mesothelin, midkine, neural cell adhesion molecule L1 (L1CAM), and transglutaminase 2 (TGM2) were measured using MAGPIX®System in plasma samples of 45 EC, 20 healthy controls and 11 patients with endometriosis. Results Significantly increased concentration in EC as compared to healthy controls were found in case of CD44 ( p <  0.001), EpCAM ( p = 0.033) and TGM2 ( p <  0.001). EpCAM and mesothelin concentrations differed based on FIGO stages. Regression analysis revealed marker panels with high accuracy in detection of EC. The highest AUC 0.937 was attributed to the 3-marker panel of CD44/TGM2/EpCAM (84% sensitivity, 100% specificity), FIGO IA samples were discriminated from more advanced stages of EC with the mesothelin/grade 1 model featuring AUC of 0.911 (95.24% sensitivity, 78.26% specificity). Conclusions Novel plasma biomarkers presenting good accuracy in diagnosing EC were found with TGM2 reported for the first time as plasma marker. It was also revealed that endometriosis may share similarities in the pattern of markers alterations characteristic for EC. Electronic supplementary material The online version of this article (10.1186/s12885-019-5556-x) contains supplementary material, which is available to authorized users.

No abstract