GEODE: antool that translatestopredictions of tuberculosis antibiotic combination efficacy.

INTRODUCTION: Tuberculosis (TB) remains the primary cause of death due to infectious disease in the world. TB, while treatable, requires an extended course of multiple antibiotics, taking 6-9 months, and many antibiotic regimens have deleterious side effects. Treatment is complicated by co-infection, emerging drug resistance, and compliance issues; accordingly, the identification of new and optimal regimens has been a recent focus. Rodent models of TB (e.g., mouse, rabbit) do not mimic some severe pathologies well, while nonhuman primate models are costly. Several computational andtools have been developed to explore drug regimen design and efficacy for TB, each providing insight into human disease dynamics.

METHODS: Here we briefly review existing tools and introduce a novel, integrated approach combiningpredictions of drug pharmacokinetics, pharmacodynamics and drug-drug interactions with a granuloma-scale computational model (). Our method capturesdynamics to test how well systematicdata predict granuloma-scale outcomes such as CFU burden and sterilization time. To evaluatemeasurements under various growth conditions and to compare to clinical and experimental datasets, we simulated five well-known regimens in our pipeline: HRZM, BPaMZ, RMZE, BPaL and HRZE.

RESULTS: We find thatmeasurements of antibiotic regimen pharmacodynamics under specific growth conditions can be used to simulate virtual granulomas consistent with low-burden human and primate granulomas.

DISCUSSION: This work provides a novel tool that can be used to quickly and efficiently evaluate drug regimens for TB.