Background: Definitive clinical trials of new chemotherapy for tuberculosis (TB) treatment require following subjects until at least six months after they discontinue taking medicine to assess durable cure, making these trials expensive and lengthy. Surrogate endpoints relating to treatment failure and relapse are currently limited to sputum microbiology, which is neither very sensitive nor very specific.
Conclusions: In this study we prospectively assessed radiographic changes using positron emission tomography (PET) and computed tomography (CT) at two (PET and CT) and six (CT only) months in a cohort of subjects with multidrug-resistant (MDR) TB who were treated with second-line TB therapy for two years and then followed up for an additional six months. CT scans were read semi-quantitatively by radiologists and computationally evaluated using custom software to provide volumetric assessment of TB-associated abnormalities. CT scans assessed by readers were predictive at six months but not two months and changes in computed abnormal volumes were predictive at both time points. Quantitative changes in FDG uptake two months after starting treatment were found to be highly predictive of long-term outcome. These radiologic markers were all more predictive than conventional sputum microbiology in distinguishing successful from unsuccessful treatment. Lesions on CT with higher radiodensity were most predictive of long-term outcome for subjects in this study. These results support the potential of radiologic biomarkers (“radiomarkers”) as possible surrogate endpoints in clinical trials of new TB drug regimens.