Klin Onkol 2024; 37(2): 102-109. DOI: 10.48095/ccko2024102.
Background: Endometrial carcinoma (EC) is the most common cancer of the female reproductive tract in developed countries. The prognosis and 5-year survival rates are closely tied to the stage diagnosis. Current routine diagnostic methods of EC are either lacking specificity or are uncomfortable, invasive and painful for the patient. As of now, the gold diagnostic standard is endometrial biopsy. Early and non-invasive diagnosis of EC requires the identification of new biomarkers of disease and a screening test applicable to routine laboratory diagnostics. The application of untargeted metabolomics combined with artificial intelligence and biostatistics tools has the potential to qualitatively and quantitatively represent the metabolome, but its introduction into routine diagnostics is currently unrealistic due to the financial, time and interpretation challenges. Fluorescence spectral analysis of body fluids utilizes autofluorescence of certain metabolites to define the composition of the metabolome under physiological conditions. Purpose: This review highlights the potential of fluorescence spectroscopy in the early detection of EC. Data obtained by three-dimensional fluorescence spectroscopy define the quantitative and qualitative composition of the complex fluorescent metabolome and are useful for identifying biochemical metabolic changes associated with endometrial carcinogenesis. Autofluorescence of biological fluids has the prospect of providing new molecular markers of EC. By integrating machine learning and artificial intelligence algorithms in the data analysis of the fluorescent metabolome, this technique has great potential to be implemented in routine laboratory diagnostics.
