BackgroundOsteoarticular tuberculosis is one of the extrapulmonary tuberculosis (EPTB) diseases, which is mainly caused by infection of Mycobacterium tuberculosis (MTB) in bone and joints. The limitation of current clinical test methods is leading to a high misdiagnosis rate and affecting the treatment and prognosis. This study aims to search serum biomarkers that can assist in the diagnosis of osteoarticular tuberculosis. MethodsProteomics can serve as an important method in the discovery of disease biomarkers. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) was used to analyze proteins in 90 serum samples, which were collected from June 2020 to December 2021, then evaluated by statistical analysis to screen potential biomarkers. After that, potential biomarkers were validated by enzyme-linked immunosorbent assay (ELISA) and diagnostic models were also established for observation of multi-index diagnostic efficacy. Results118 differential expressed proteins (DEPs) were obtained in serum after statistical analysis. After the diagnostic efficacy evaluation and clinical verification, inter-alpha-trypsin inhibitor heavy chain H2 (ITIH2), complement factor H-related protein 2 (CFHR2), complement factor H-related protein 3 (CFHR3), and complement factor H-related protein 5 (CFHR5) were found as potential biomarkers, with 0.7167 (95 %CI: 0.5846–0.8487), 0.8600 (95 %CI: 0.7701–0.9499), 0.8150 (95 %CI: 0.6998–0.9302), and 0.9978 (95 %CI: 0.9918–1.0040) AUC value, respectively. The remaining DEPs except CFHR5 were constructed as diagnostic models, the diagnostic model contained CFHR2 and CFHR3 had good diagnostic efficacy with 0.942 (95 %CI: 0.872–0.980) AUC value compared to other models. ConclusionThis study provides a reference for the discovery of serum protein markers for osteoarticular tuberculosis diagnosis, and the screened DEPs can also provide directions for subsequent pathogenesis research.
Read full abstract