Exploring novel salivary host biomarkers for immunological diagnosis of tuberculosis: A preliminary biomarker discovery study

Tuberculosis is a serious public health concern on a global scale, which emphasises the critical need for quick and precise diagnostic and treatment response monitoring techniques. In this study, Luminex multiplex immunoassay was used to detect the concentrations of 37 host biomarkers in saliva samples from 46 patients newly diagnosed with active pulmonary tuberculosis (PTB) and 46 patients with other respiratory diseases (ORD). Multiple logistic regression and the area under the receiver operator characteristics curve (AUC) were used to evaluate the diagnostic accuracy of biomarkers, which showed significant differences between the 2 groups. This study reported that Fractalkine exhibited the highest diagnostic accuracy and excellent discriminatory power, with statistically significant results (p ≤ 0.05), an AUC of 0.91, 89.1 % sensitivity and 76.1 % specificity, highlighting its strong potential to distinguish PTB cases from ORD cases. Additionally, our study found that the median levels of IL-17A, IL-23, and VEGF were statistically significant (p ≤ 0.05). General discriminant analysis further identified Fractalkine, VEGF, GM-CSF, IL-23, and IL-1α as the top five most effective biomarkers for combinations. The backward elimination approach demonstrated the potential usefulness of a four-marker combination (Fractalkine + GM-CSF + IL-23 + IL-1α) as a confirmatory diagnostic tool by achieving the greatest overall diagnostic accuracy with an AUC of 0.94 and 91.3 % specificity. Thus, combining multiple markers with high discriminating power may improve diagnostic performance and subsequently provide a more accurate, non-invasive saliva-based PTB diagnostic tool.