Tuberculosis (TB) is a leading global public health problem. To achieve the end TB strategy, non-invasive markers for diagnosis and treatment monitoring of TB disease are urgently needed, especially in high-endemic countries such as China. Interferon-gamma release assays (IGRAs) and tuberculin skin test (TST), frequently used immunological methods for TB detection, are intrinsically unable to discriminate active tuberculosis (ATB) from latent tuberculosis infection (LTBI). Thus, the specificity of these methods in the diagnosis of ATB is dependent upon the local prevalence of LTBI. The pathogen-detecting methods such as acid-fast staining and culture, all have limitations in clinical application. ImmunoScore (IS) is a new promising prognostic tool which was commonly used in tumor. However, the importance of host immunity has also been demonstrated in TB pathogenesis, which implies the possibility of using IS model for ATB diagnosis and therapy monitoring. In the present study, we focused on the performance of IS model in the differentiation between ATB and LTBI and in treatment monitoring of TB disease. We have totally screened five immunological markers (four non-specific markers and one TB-specific marker) and successfully established IS model by using Lasso logistic regression analysis. As expected, the IS model can effectively distinguish ATB from LTBI (with a sensitivity of 95.7% and a specificity of 92.1%) and also has potential value in the treatment monitoring of TB disease.
Keywords: ATB; ImmunoScore; LTBI; diagnosis; therapy monitoring.