Urine supernatant reveals a signature that predicts survival in clear‐cell renal cell carcinoma

• Published in: BJU international Vol. 132; no. 1; pp. 75 - 83
• Main Authors: Daza, Jorge, Salomé, Bérengère, Okhawere, Kennedy, Bane, Octavia, Meilika, Kirolos N., Korn, Talia G., Qi, Jingjing, Xe, Hui, Patel, Manishkumar, Brody, Rachel, Kim‐Schulze, Seunghee, Sfakianos, John P., Lewis, Sara, Rich, Jordan M., Zuluaga, Laura, Badani, Ketan K., Horowitz, Amir
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 01.07.2023
• Subjects: Biomarkers; Cell survival; Genomes; Inflammation; kcsm; Kidney cancer; KidneyCancer; Medical prognosis; molecular biomarkers; Nomograms; outcomes; Proteins; Proteomics; Renal cell carcinoma; Risk factors; risk group; Risk groups; Transcriptomics; tumour microenvironment; Urine; uroonc; kcsm; kidney cancer; uroonc; KidneyCancer; inflammation; outcomes; molecular biomarkers; tumour microenvironment; risk group
• ISSN: 1464-4096; 1464-410X; 1464-410X
• DOI: 10.1111/bju.15989

Objective To profile the cell‐free urine supernatant and plasma of a small cohort of clear‐cell renal cell carcinoma (ccRCC) patients by measuring the relative concentrations of 92 proteins related to inflammation. Using The Cancer Genome Atlas (TCGA), we then performed a targeted mRNA analysis of genes encoding the above proteins and defined their effects on overall survival (OS). Subjects/patients and Methods Samples were collected prospectively from ccRCC patients. A multiplex proximity extension assay was used to measure the concentrations of 92 inflammation‐related proteins in cell‐free urine supernatants and plasma. Transcriptomic and clinical information from ccRCC patients was obtained from TCGA. Unsupervised clustering and differential protein expression analyses were performed on protein concentration data. Targeted mRNA analysis on genes encoding significant differentially expressed proteins was performed using TCGA. Backward stepwise regression analyses were used to build a nomogram. The performance of the nomogram and clinical benefit was assessed by discrimination and calibration, and a decision curve analysis, respectively. Results Unsupervised clustering analysis revealed inflammatory signatures in the cell‐free urine supernatant of ccRCC patients. Backward stepwise regressions using TCGA data identified transcriptomic risk factors and risk groups associated with OS. A nomogram to predict 2‐year and 5‐year OS was developed using these risk factors. The decision curve analysis showed that our model was associated with a net benefit improvement compared to the treat‐all/none strategies. Conclusion We defined four novel biomarkers using proteomic and transcriptomic data that distinguish severity of prognosis in ccRCC. We showed that these biomarkers can be used in a model to predict 2‐year and 5‐year OS in ccRCC across different tumour stages. This type of analysis, if validated in the future, provides non‐invasive prognostic information that could inform either management or surveillance strategies for patients.