Liq_ccRCC: Identification of Clear Cell Renal Cell Carcinoma Based on the Integration of Clinical Liquid Indices

• Main Authors: Jianhong Zhao, Jiangpeng Wu, Jinyan Wei, Xiaolu Su, Yanjun Chai, Shuyan Li, Zhiping Wang
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2021-01-01
• Series: Frontiers in Oncology
• Subjects: Liq_ccRCC; clear cell renal cell carcinoma; subtype differentiation; liquid indices; machine learning
• Online Access: https://www.frontiersin.org/articles/10.3389/fonc.2020.605769/full

Currently, preoperative diagnosis and differentiation of renal clear cell carcinoma and other subtypes remain a serious challenge for doctors. The liquid biopsy technique and artificial intelligence have inspired the pursuit of distinguishing clear cell renal cell carcinoma using clinically available test data. In this work, a method called liq_ccRCC based on the integration of clinical blood and urine indices through machine learning approaches was successfully designed to achieve this goal. Clinically available biochemical blood data and urine indices were collected from 306 patients with renal cell carcinoma. Finally, the integration of 18 top-ranked clinical liquid indices (13 blood samples and 5 urine samples) was proven to be able to distinguish renal clear cell carcinoma from other subtypes of renal carcinoma by cross-valuation with an AUC of 0.9372. The successful introduction of this identification method suggests that subtype differentiation of renal cell carcinoma can be accomplished based on clinical liquid test data, which is noninvasive and easy to perform. It has huge potential to be developed as a promising innovation strategy for preoperative subtype differentiation of renal cell carcinoma with the advantages of convenience and real-time testing. liq_ccRCC is available online for the free test of readers at http://lishuyan.lzu.edu.cn/liq_ccRCC.