In Search of Relevant Urinary Biomarkers for Thyroid Papillary Carcinoma and Benign Thyroid Nodule Differentiation, Targeting Metabolic Profiles and Pathways via UHPLC-QTOF-ESI+-MS Analysis.

Background: Identification of specific urine metabolic profiles for patients diagnosed with papillary thyroid carcinoma (TC) vs. benign nodules (B) to identify specific biomarkers and altered pathways compared to those of healthy controls (C).

Methods: Patient urine samples were collected, before surgery and after a histological confirmation of TC (n = 30) and B (n = 30), in parallel with sample collection from healthy controls (n = 20). The untargeted and semi-targeted metabolomic protocols were applied using UPLC-QTOF-ESI⁺-MS analysis, and the statistical analysis was performed using the Metaboanalyst 6.0 platform. The results for the blood biomarkers, previously published, were compared with the data obtained from urine sampling using the Venny algorithm and multivariate statistics.

Results: Partial least squares discrimination, including VIP values, random forest graphs, and heatmaps (p < 0.05), together with biomarker analysis (AUROC ranking) and pathway analysis, suggested a specific model for the urinary metabolic profile and pathway alterations in TC and B vs. C, based on 190 identified metabolites in urine that were compared with the serum metabolites. By semi-targeted metabolomics, 10 classes of metabolites, considered putative biomarkers, were found to be responsible for specific alterations in the metabolic pathways, from polar molecules to lipids. Specific biomarkers for discrimination were identified in each class of metabolites that were either upregulated or downregulated when compared to those of the controls.

Conclusions: The lipidomic window was the most relevant for identifying biomarkers related to thyroid cancer and benign conditions, since this study detected a stronger involvement of lipids and selenium-related molecules for metabolic discrimination.