Cure-screening predicts mechanism of post-antibiotic relapse in Tuberculosis

Abstract

Tuberculosis (TB) remains a global health concern, as Mycobacterium tuberculosis (Mtb) currently infects a quarter of the worlds population. Though many TB patients sterilize infection with short-course treatment, regimens shorter than 4 months risk post-treatment, potentially facilitating drug resistance. Granulomas are spatially-heterogeneous hallmark immune structures that form during TB, comprising immune cells and caseum, necrotic tissue that can trap Mtb in a non-replicating state. Two mechanisms causing relapse have been hypothesized: persistence, where treatment kills all replicating Mtb, and relapse follows once Mtb trapped within caseum returns to a replicative niche; and threshold, where replicating Mtb remain alive below detectable levels. Typically, clinical relapse is described as TB recurrence <2 years after a misdiagnosis of cure upon treatment completion (MDxC). Comparatively, many experimental models cannot screen for cure and examine relapse [~]2-months after treatment completion. Capacity to untangle these considerations in vivo are limited. Here, we examine the impact of study design (e.g., cure screening) on mechanisms underpinning reported instances of relapse using our computational model capturing whole-host Mtb infection dynamics, HostSim. Simulations uncover rates of reported relapse depending on whether hosts are screened for cure upon treatment completion. If not screened, then relapse is likely driven by incomplete sterilization of replicating bacteria; whereas cure-screened relapse is most likely to be caused by gradual expansion of non-replicating Mtb from within caseum into cellular areas of a granuloma. This suggests that TB patients that relapse after MDxC may best be treated with caseum-penetrating antibiotics such as rifamycin-class antibiotics. ImportanceIncomplete treatment of TB leads to risks relapse, which may occur years later. The threat of relapse is the main reason TB treatment takes 4-9 months. Predictors of relapse are not well-defined given variability in technical definitions of relapse and nuances of study designs. Here, we simulate both clinical and experimental relapse studies, including multiple diagnostic tests and relapse definitions, using biologically-based computation. We find that two hypothesized types of relapse that each potentially require a different treatment strategy are simultaneously at play. Simulations suggest that relapse after a "cure" diagnosis (most clinical studies) is caused by reactivation of non-replicating bacteria hidden from treatment within necrotic granuloma tissue or other sites, whereas experiments that cannot test for cure post-treatment are likely to report relapse caused by incomplete sterilization of replicating Mtb. Predictions depend on the currently unclear relationship between bacterial burden and clinical symptoms.