Exploring immunological host biosignatures in saliva: A novel framework for early diagnosis and treatment response...

Tuberculosis (TB) emphasises the critical need for quick and precise diagnosis techniques as well as treatment response monitoring, as it has become a major global source of disease and mortality. In this present study, saliva samples from 46 patients with newly diagnosed active pulmonary tuberculosis (PTB), 46 patients with other respiratory disorders (ORD) and 46 Healthy Individuals (HI) were collected and evaluated for the concentrations of 37 host biomarkers using Luminex multiplex immunoassay. Significant variations between groups have been identified in the diagnostic accuracy of biomarkers as assessed by multiple logistic regression and area under curve (AUC). Fractalkine exhibited the greatest diagnostic accuracy and discriminatory capability, with high sensitivity, and specificity, highlighting its effectiveness as the best biomarker for differentiating PTB patients from ORD cases and HI. The five-marker combination (Fractalkine + VEGF + IL-23+IL-17A + GM-CSF) achieved the best diagnostic performance, with an AUC of 0.98, accuracy of 90.2 % suggesting a strong ability to identify PTB cases from ORD. In addition, the five-marker panel (Fractalkine + IL-1α+ GRO + IP-10+MCP-1) achieved excellent performance with an AUC of 0.99 and an accuracy of 92.4 %, for distinguishing PTB from HI. The salivary concentrations of Fractalkine, IL-17A, IL-23, and VEGF decreased significantly at 2 and 6 months, indicating their potential as biomarkers for monitoring TB treatment responses. These findings provide valuable insights into non-invasive TB diagnostics and treatment monitoring, warranting further validation in a larger cohort to facilitate clinical translation.