Work

Identificación de biomarcadores urinarios de cáncer vesical mediante análisis metabolómico por espectrometría de masas

Year: 2020
Type: dissertation
Study design: Observational Study
Abstract: INTRODUCTION: Bladder cancer (BC) has a great epidemiological, economic and clinical relevance. Cystoscopy, an invasive technique, continue to be the standard methods of diagnosis and follow-up, along with cytology, which has a high specificity but very limited sensitivity in low-risk tumors. The identification of new highly sensitive and specific biomarkers (bioM) would reduce cystoscopies, as well as their morbidity and it would improve patients’ quality of life. Nowadays, there are different proteomic and genomic bioM that have shown greater sensitivity than cytology, but their clinical applications have not allowed since their limited specificity. Metabolomic systematically studies the concentrations and flows of all metabolites present in a biological system reflecting the phenotypic state of an organism in a dynamic way. Hence, improving metabolomic analysis techniques, would facilitate the identification of metabolic profiles associated with metabolic reprogramming of tumor cells in several cancers. OBJECTIVE: To identify a specific urinary metabolic footprint of BC and study its value as a tumor bioM for diagnosis-detection of disease and monitoring the BC recurrence during follow-up after bladder transurethral resection (TURB). MATERIAL AND METHOD: During the period between February 2012 and December 2015, non-muscle invasive bladder tumor (NMIBC) patients were selected who underwent to TURB. A protocol was designed to collect the initial urine sample before the TURB (BC), and a second sample after the TURB (CTRL control). Once the urine samples were prepared, the analysis was carried out by means of high-performance liquid chromatography in reverse phase and mass spectrometry (UPLC-TOF-MS). Subsequently, the discriminant analysis was monitored by Partial Least Square Discriminat Analysis (PLS-DA), which allowed us to find differences between the metabolic profiles of the BC and CTRL samples. RESULTS: 53 patients with NMIBC are included in the study. The metabolomic study comprised the analysis of a total of 244 chromatograms: 129 urine samples (BC+ CTRL) and 55 follow up quarterly monitoring urine samples (MNT), in addition to 45 Quality Control (QC) samples, 3 control blanks samples and a sub-set of duplicate samples (12 samples). The urine samples were divided into two subsets that were used for the development of a discriminant multivariate model (calibration set) and for the evaluation of the predictive capacity of that model in future samples (validation set). Through the first analysis, the PLS-DA model developed with the calibration samples select a reduced set of 108 highly discriminating metabolic variables (p-values <0.05). The second PLS-DA model, using the 108 selected variables and the 81 samples of the validation subset, classify correctly 36 of the 43 BC samples and 25 of the 38 CTRL, with a diagnostic accuracy of 75% and an estimated AUROC value of 0.77, with a sensitivity of 83% (95% CI, 69-93%) and a specificity of 66% (95% CI, 49-80%). CONCLUSIONS: The UPLC-TOF-MS analysis of urine samples collected through before and after TURB design, identify a metabolic footprint of NMIBC, demonstrating that the results are associated with the presence of the macroscopic BC and that they are generated after TURB.
Author: Trassierra Villa, Marta
Language: English
Open Access status: green
OpenAlex ID: W7038189272
Cites: 134
Cited by:
Related to: 9