Background Permanent lung injury and impaired function are common despite TB cure. Host-directed anti-inflammatory therapies may prevent this injury. Early biomarkers of lung inflammation and function can facilitate their evaluation.

Methods In an ongoing study supported by the Bill and Melinda Gates Foundation, HIV-uninfected patients with radiographically moderately or far advanced sputum smear-positive pulmonary tuberculosis receive rifabutin-substituted standard therapy plus either CC-11050 (phosphodiesterase inhibitor), everolimus (mTOR inhibitor), auranofin (gold salt), cholecalciferol, or control, during months 1–4. Study leadership is blinded as to assigned treatments. 18F-fluorodeoxyglucose positron emision tomography (PET) and computed tomography (CT) are performed at baseline and at week 8. Total lung glycolytic activity (SUVbw*ml) and radiodensity (modified HU*ml) are measured using MIM software. Sputum culture, spirometry, 6 min walk test (6MWT), and other biomarkers are performed at multiple time points. Follow-up continues to month 18. This analysis includes only baseline and week 8 data.

Results Presently, 160/200 participants are enrolled. At baseline, patients have a high burden of infection (median time to detection [TTD] in automated liquid culture 5 days). Median baseline FEV1% of predicted (63%) and 6MWT (402 meters) are typical of moderate to severe chronic lung disease. Baseline TTD, PET, CT, FEV1% and 6MWT are all highly correlated (median rank test p=0.0018). All 5 parameters changed significantly during 8 weeks of treatment (p<0.001). Analysis of adjusted log change from baseline shows PET and CT remain highly correlated (p<0.001), and weakly correlated with FEV1% and 6MWT. TTD shows no correlation with any other endpoint.

Conclusion Quantitative markers of infection, inflammation, and function are markedly abnormal and highly correlated at baseline in patients with pulmonary tuberculosis. Quantitative CT may substitute for PET as a more readily-performed measure of lung inflammation. The dissociation of microbiologic responses from inflammation and function supports a role for HDTs in TB.