Blood transcriptomic signatures predict poor treatment outcomes in drug-susceptible pulmonary TB in Brazil.

Background: Non-sputum biomarkers to monitor tuberculosis (TB) treatment and predict poor outcomes are lacking. We evaluated host-blood transcriptomic signatures for treatment monitoring and prognosis (death, treatment failure, recurrence) in adults with pulmonary TB. Methods: Adults with culture-confirmed, drug-susceptible pulmonary TB were enrolled at five Brazilian sites. Whole-blood PAXgene samples were collected at baseline, month 2 (M2), and end of treatment (EoT). Treatment failure was defined as sputum culture positivity at month 5 or later. Participants were followed for 24 months from treatment initiation for clinical or microbiological TB recurrence. Unfavourable outcomes were matched ~1:3 to recurrence-free cure. Twenty-two published blood transcriptomic signatures were measured by microfluidic RT-qPCR and benchmarked against the WHO Target Product Profile (TPP) criteria. Results: We matched 263 participants with recurrence-free cure to 33 with treatment failure, 24 who died (TB/unknown cause), and 9 with recurrence. Signature scores generally declined from baseline to EoT. Multiple signatures predicted recurrence at baseline and M2 (AUC range 0.71-0.91), with waning performance at EoT (AUC range 0.42-0.89). Against the WHO TPP, 2/22 signatures met minimum criteria at baseline, 13/22 at M2, and none at EoT. Prediction of treatment failure was poor across timepoints (AUC <0.70). In contrast, Thompson5 and others at baseline predicted death during treatment or follow-up (AUC ≥0.80). Conclusion: Blood transcriptomic signatures tracked treatment response and predicted recurrence and death, meeting WHO TPP benchmarks at baseline and M2. These findings support prospective, biomarker-guided trials to individualise TB therapy-shortening regimens for early responders and intensifying care for high-risk patients. Funding: This work was supported by the U.S. National Institutes of Health, CRDF Global, and Departamento de Ciência e Tecnologia (DECIT) - Secretaria de Ciência e Tecnologia (SCTIE), Ministério da Saúde (MS), Brazil.