Li, Taixian, Zhang, Yanqiong, Wang, Rongtian, Xue, Zhipeng, Li, Shangzhu, Cao, Yuju, Liu, Daobing, Niu, Yanfang, Mao, Xia, Wang, Xiaoyue, Li, Weijie, Guo, Qiuyan, Guo, Minqun, Lin, Na, and Chen, Weiheng
Abstract Steroid-induced osteonecrosis of the femoral head (SONFH) is difficult to be diagnosed at the early stages when it can be administrated effectively. Yet, to date no study has been performed to identify diagnostic biomarkers and to develop diagnostic models for SONFH. In the current study, a total of 60 SONFH patients with Association Research Circulation Osseous (ARCO) stages I–IV, and 20 controls were enrolled and divided into the discovery and validation cohorts. The serum samples were collected and the gene expression profiles were detected by microarray analysis based on the discovery cohort. Then, eight genes (BIRC3, CBL, CCR5, LYN, PAK1, PTEN, RAF1 and TLR4) were identified as the candidate serum biomarkers of SONFH due to the significant differential expression patterns and the topological importance in the interaction network of SONFH-related differentially expressed genes. Functionally, these candidate serum biomarkers were significantly involved into several pathological processes during SONFH progression, such as the immune regulation and inflammation, bone metabolism and angiogenesis. After that, a prediction model for the diagnosis of SONFH was constructed using Partial least squares regression based on the serum levels of the candidate biomarkers. Notably, both the 10-fold cross-validation and the independent dataset test demonstrated the good performance of this model. In conclusion, our study discovered eight promising serum biomarkers and developed the multi-biomarker-based prediction model as a new, potential and non-invasive diagnostic tool for the detection of SONFH, as well as benefit the administration of SONFH in a daily clinical setting. Graphical abstract Unlabelled Image Highlights • Eight serum gene biomarkers were identified for the detection of SONFH. • The eight-biomarker serum gene signature-based diagnostic model for SONFH was developed. • The eight-biomarker serum gene signature-based diagnostic model displayed the good performance in predicting SONFH. [ABSTRACT FROM AUTHOR]