Re-evaluating progression and pathways following Mycobacteria tuberculosis infection within the spectrum of tuberculosis disease

BackgroundTraditional understanding of the risk of progression from Mycobacterium tuberculosis (Mtb) infection to tuberculosis (TB) disease overlooks nuance across a spectrum of disease. MethodsWe developed a deterministic model of Mtb infection and minimal (pathological damage but not infectious), subclinical (infectious but no reported symptoms), and clinical (infectious and symptomatic) TB disease, informed by a rigorous evaluation of data from a systematic review of TB natural history. Using a Bayesian approach, we calibrated the model to data from historical cohorts that followed tuberculin-negative individuals to tuberculin conversion and TB disease, as well as data from cohorts that followed progression and regression between disease states, disease state prevalence ratios, disease duration, and mortality. We estimated incidence, pathways, and ten-year outcomes following Mtb infection for a simulated cohort. Results90.8% (95% uncertainty interval, UI, 90.2-91.3) of individuals self-cleared within 10 years of infection, while 9.3% (95% UI 8.4-10.0) progressed to TB disease. Of those, 68.1% (95% UI 65.1-71.1) developed infectious disease, and 32.7% (95% UI 29.7-35.7) progressed to clinical disease. While 93% of progression to minimal disease occurred within two years of infection, only 63% and 38% of subclinical and clinical disease, respectively, occurred within this period. Multiple progression pathways from infection were necessary to calibrate the model, and 48.8% (95% UI 45.0-52.6) of those who developed infectious disease undulated between disease states. ConclusionsWe identified highly heterogeneous pathways across disease states after Mtb infection, highlighting the need for clearly defined disease thresholds to inform more effective prevention and treatment efforts to end TB.