Background and Objectives Point-of-care testing using nonsputum samples like serum or plasma proteins can improve tuberculosis (TB) patients access to a definitive diagnosis, especially in resource-constrained and remote areas. Recently, approximately 400 proteins were identified as playing a role in the pathogenesis of TB, offering a translational clinical research repository for TB. In a previous manuscript, we proved the potential use of these proteins for point-of-care testing for active TB diagnosis. The present work aims to confirm the performance of single and combination proteins to select the best candidate biomarkers for further development as a diagnostic testing tool for active TB. Methods Seventy-four participants were assessed on the diagnostic performance of 17 single proteins and combinations of 2 to 4 proteins to diagnose active TB. The selection criteria included differential expression of the proteins between active TB and community-acquired pneumonia (CAP) and a performance rate ≥70% for active TB. Results SULT4A1, WASPF3, SPTLC1, FAM107B, SORCS2, and CYTOb561 were differentially expressed in TB compared to CAP patients. Two single proteins, SULT4A1 and WASPF3, performed ≥70% to discriminate active TB from CAP patients. The diagnostic performance of 3 protein-based combinations of active TB was 81% after leave-one-out cross-validation. Conclusion Single proteins and 3 protein-based combinations are candidate biomarkers for diagnosing active TB disease. A large and prospective study will confirm their performance as complementary diagnostic tools to rapid diagnostic methods for detecting active TB.
No takes yet. Share an insight, caveat, or question.
Essone et al. (2024) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: