P007

1:Sciensano, Brussels, Belgium; 2:University of Antwerp; 3:Univ. Lille, CNRS, Inserm, CHU Lille, Institut Pasteur de Lille; 4:Institut Pasteur, Université Paris Cité; 5:Institute of Tropical Medicine, Antwerp, Belgium

The growing challenge of antibiotic resistance in Mycobacterium tuberculosis emphasizes the need for rapid phenotypic drug susceptibility testing (pDST). While DNA-based methods have greatly improved TB diagnostics worldwide, they fundamentally depend on prior pDST data. Moreover, as new drugs are developed and resistance mutations continue to emerge, pDST remains indispensable. Although MGIT is a reliable standard for phenotypic testing of many anti-TB drugs, it is time-consuming and known to miss certain resistant phenotypes. We present TRACeR-TB, a fast, RNA-based pDST method that distinguishes resistant from susceptible strains by measuring transcriptional stress responses triggered by a brief, 6h antibiotic exposure. Using multiplex panels of 8-9 mRNA biomarkers per drug, TRACeR-TB enables accurate resistance profiling directly from crude lysates. Validation on 238 clinical isolates showed higher sensitivity compared to MGIT, with improved performance in detecting borderline-resistant strains, particularly for rifampicin and bedaquiline. TRACeR-TB demonstrated full concordance with MGIT for drugs with reliable MGIT results (moxifloxacin and isoniazid), and proved effective in identifying resistant phenotypes undetected by genotypic methods. Furthermore, the broad applicability of the assay is illustrated through its use in testing novel drug-boosting compounds and its ability to detect intracellular drug-inflicted stress levels, enabling hit-to-lead optimization in and ex vivo. As a rapid and mutation-independent approach, TRACeR-TB serves not only as a valuable diagnostic tool but also as a platform for investigating resistance mechanisms and supporting the development of novel therapeutic strategies, offering a complementary DST to bridge existing gaps in TB diagnostics and research.