Abstract
Multidrug-resistant tuberculosis (MDR-TB), defined as tuberculosis (TB) resistant to at least isoniazid and rifampicin, is a major concern of TB control worldwide. However, the diagnosis of MDR-TB remains a huge challenge to its prevention and control. To identify new diagnostic methods for MDR-TB, a mass spectrometry strategy of data-independent acquisition and parallel reaction monitoring was used to detect and validate differential serum proteins. The bioinformatic analysis showed that the functions of differential serum proteins between the MDR-TB group and the drug-sensitive tuberculosis group were significantly correlated to the complement coagulation cascade, surface adhesion and extracellular matrix receptor interaction, suggesting a disorder of coagulation in TB. Here, we identified three potential candidate biomarkers such as sCD14, PGLYRP2 and FGA, and established a diagnostic model using these three candidate biomarkers with a sensitivity of 81.2%, a specificity of 90% and the area under the curve value of 0.934 in receiver operation characteristics curve to diagnose MDR-TB. Our study has paved the way for a novel method to diagnose MDR-TB and may contribute to elucidate the mechanisms underlying MDR-TB.