Baseline host inflammatory and transcriptional profiles associated with structural and functional recovery in drug-resistant tuberculosis

BackgroundIn extensively drug-resistant and pre-extensively drug-resistant TB, bacteriology-based monitoring often fails to capture structural lung recovery and patient-reported functional health. We aimed to characterize multidomain treatment response and examine host inflammatory and transcriptional features associated with incomplete recovery. MethodsWe conducted an ancillary analysis of a prospective, open-label, pilot study evaluating adjunctive ibuprofen in XDR-TB (NCT02781909). Participants were assessed at baseline and during treatment using TBS, chest radiography, sputum culture, SGRQ, blood cell indices, plasma cytokines, and whole-blood transcriptomic profiling. Clinical and laboratory measures were compared across outcome groups, and blood transcriptional profiles were analyzed in relation to treatment outcomes. ResultsHere we show that microbiological and symptomatic improvement occurred earlier than radiological and functional recovery. Higher baseline systemic inflammation, including elevated NLR, SII, and IL-6, as well as increased expression of interferon-related genes such as CD274 and GBP5, were associated with poorer radiological and SGRQ outcomes at 6 months. In contrast, transient elevations of IL-8 and IL-4 were associated with early bacteriological clearance. IL-8 was the only plasma biomarker consistently correlated with symptom severity, radiological findings, and functional health. ConclusionsTreatment response in drug-resistant TB is asynchronous across biological domains. Integrated host profiling identifies inflammatory and transcriptional features associated with incomplete structural and functional recovery, supporting the use of multidimensional endpoints to better capture long-term outcomes and inform individualized patient management. Plain Language SummaryPeople with highly drug-resistant tuberculosis can clear the infection but still experience lung damage and reduced quality of life after treatment. In this study, we examined recovery using several measures, including symptoms, chest X-rays, blood markers of inflammation, and gene activity, in addition to tests for tuberculosis bacteria. We analyzed data and stored samples from a small clinical trial to see how these measures changed over time. We found that lung structure and quality of life improved more slowly than bacterial clearance. People with higher levels of inflammation before treatment were more likely to have ongoing lung changes and poorer quality of life later. These results suggest that tuberculosis care should look beyond bacterial clearance and include monitoring inflammation to better support long-term recovery.