RESUMO
Cancer is responsible for millions of deaths worldwide, but most base diseases may be cured if detected early. Screening tests may be used to identify early-stage malignant neoplasms. However, the major screening tool for prostate cancer, the prostate-specific antigen test, has unsuitable sensitivity. Since cancer cells may affect the pattern of consumption and excretion of nucleosides, such biomolecules are putative biomarkers that can be used for diagnosis and treatment evaluation. Using a previously validated method for the analysis of nucleosides in blood serum by capillary electrophoresis with UV-vis spectroscopy detection, we investigated 60 samples from healthy individuals and 42 samples from prostate cancer patients. The concentrations of nucleosides in both groups were compared and a multivariate partial least squares-discriminant analysis classification model was optimized for prediction of prostate cancer. The validation of the model with an independent sample set resulted in the correct classification of 82.4% of the samples, with sensitivity of 90.5% and specificity of 76.7%. A significant downregulation of 5-methyluridine and inosine was observed, which can be indicative of the carcinogenic process. Therefore, such analytes are potential candidates for prostate cancer screening. Graphical Abstract Separation of the studied nucleosides and the internal standard 8-Bromoguanosine by CE-UV (a); classification of the external validation samples (30 from healthy volunteers and 21 from prostate cancer patients) by the developed Partial Least Square - Discriminant Analysis (PLS-DA) model with accuracy of 82.4% (b); Receiver Operating Characteristics (ROC) curve (c); and Variable Importance in the Projection (VIP) values for the studied nucleosides (d). A significant down-regulation of 5- methyluridine (5mU) and inosine (I) was observed, which can be indicative of the presence of prostate tumors.
Assuntos
Eletroforese Capilar/métodos , Nucleosídeos/sangue , Neoplasias da Próstata/diagnóstico , Espectrofotometria Ultravioleta/métodos , Biomarcadores Tumorais , Humanos , Masculino , Estrutura Molecular , Nucleosídeos/química , Nucleosídeos/metabolismoRESUMO
The development and validation of methodologies for the analysis of biological samples is of outcome importance in order to obtain trustworthy results. This work reports a novel CE-UV method for the assessment of nucleosides, putative tumor biomarkers, in blood serum. The separation of seven nucleosides within c.a. 20 min has been achieved with: BGE 30 mmol/L borate at pH 9.90, 50 mmol/L CTAB, and 10% methanol; V = -10 kV; T = 20°C; and capillary dimensions of 56 cm × 50 µm. The sample plug was concentrated by a modified large volume sample stacking strategy that provided better detectability. Validation showed that the method is suitable for bioanalytical purposes and initial applications in serum samples from healthy subjects are also presented. Finally, statistical methods were applied to verify the effect of characteristics such as age, smoking habits, and alcohol consumption on nucleoside concentrations in blood serum. Univariate statistical analysis tests emphasized the need for age matching, which was confirmed by PCA-DA and PLS-DA. Cancer history in the nearby family may also interfere in nucleoside levels in blood serum, since adenosine concentrations were statistically higher for volunteers who declared having diseased relatives.