Project Summary
The long duration and multidrug nature of tuberculosis (TB) treatment regimens pose obstacles to treatment
completion for patients and providers, but also challenge research efforts to develop new and improved regi-
mens. Without a validated early biomarker able to discriminate regimens with different curative potential, the field
depends heavily on preclinical models. However, TB regimen development has been slow, often hindered and
misled by poor preclinical-to-clinical translation. Animal models and pharmacometric modeling played key roles
in developing groundbreaking novel regimens capable of shorting treatment durations for drug-susceptible and
multidrug-resistant TB to 4 and 6 months, respectively. However, these tools are still not utilized to their full
potential. Recent successes in TB drug discovery have produced additional clinical and preclinical drug candi-
dates, which have revealed a new challenge: with many possible combinations, how to prioritize multidrug regi-
mens to test in resource-intensive clinical trials? The most efficient “critical path” of preclinical experiments and
modeling to follow to identifying and optimizing the best regimens for advancement to clinical trials remains
uncertain. The proposed project will establish a comprehensive, collaborative, multidisciplinary consortium of
scientific leaders, drug developers and other stakeholders to develop and pursue a preclinical and translational
research agenda to identify novel regimens with the greatest potential for clinical success in adults and children
with TB. The overarching hypothesis is that integration and analysis of specific preclinical and early clinical data
using validated models and tools will enable data-driven clinical trial simulations that yield quantitative predictions
of Phase 2 and Phase 3 trial endpoints useful to identify and rank regimens with the highest probability of clinical
success in patients across the age and disease spectrum of pulmonary TB. Initial efforts will be aimed at refining
and validating preclinical in vivo (BALB/c and C3HeB/FeJ mouse) and in vitro models and translational tools by
leveraging the largest data warehouse on TB drugs and regimens ever assembled, with data spanning from
early preclinical stages to Phase 3 clinical trials, for both back translation and forward prediction approaches to
validation. The overall goal is to develop a fully data- and knowledge-driven approach to evaluate, prioritize and
optimize novel drug regimens for clinical trials requiring only preclinical and early clinical trial data. In addition to
endpoints based on bacterial burden and relapse (in mice), the RS ratio, a new portable biomarker of bacterial
“health”, will be evaluated as a complementary pharmacodynamic (PD) biomarker to increase efficiency and
predictive accuracy of preclinical studies. The expected outcomes are novel methodologies of combining pre-
clinical and early clinical data, a defined set of critical path experiments with predictive value, and a framework
for model-informed decision-making based on quantitative predictions of clinical outcomes for emerging regi-
mens prior to initiation of phase 2/3 trials. The predictions will be used to rank order candidate regimens for
advancement to clinical trials.