Table 4. Research priorities and challenges for investigating subclinical and non-infectious TB.

Research Gap	Notes/challenges
Diagnosis
Reference standards for new TB states	Validated reference standards need to be developed for subclinical TB and Mtb infection.
Validated operational tests for viable Mtb	No current tools exist to confirm the presence of Mtb in the body when routine microbiological tests are negative (e.g. on sputum for pulmonary TB or biopsies for extrapulmonary TB).
Validated operational tests for host response	A confirmatory biomarker is needed to confirm whether pathology that is evident on imaging is due to a currently active disease process. It is possible that that some biomarkers will be unique across states, reflective of current disease processes (i.e. blood RNA), and some will be detectable across multiple states (i.e. Mtb-specific T cell activation).
Validated operational tests for macroscopic pathology	Digital CXR and associated CAD technologies can detect evidence of macroscopic pathology but not with the sensitivity of more high-resolution cross-sectional imaging such as CT which are not practical operational tools. Developments in CXR technology or improvement in AI-based CAD software are needed to facilitate improved detection of macroscopic pathology. No imaging modality i specific for TB pathology hence radiotracers or additional biomarkers should be developed to detect to provide this specificity. Ultimately a diagnostic test capturing the host response associated with macroscopic pathology may omi the need for imaging.
Validated operational tests for infectiousness	Need to establish the performance of new testing modalities (e.g. bio-aerosol, tongue swab) in detecting and quantifying the degree to which index cases can cause new Mtb infections, as compared to current sputum-based tests.
Standardised protocol to detect symptoms & signs of TB	Limitations of current thresholds (coush ≥ 2 weeks, WHO 4 symptom screen [any of current cough, fever, night sweats, or weight loss]) are known, but need empirically informed alternatives that provide workable balance between sensitivity and specificity. Need to consider which symptoms and signs to include, their duration and severity.
TB screening strategies	To use tools to measure all three dimensions of disease and report the tests used, to determine sensitivity of approach. Targeting high risk populations and piggybacking or embedding subclinical and non-infectious TB screening activities into other activities can improve TB detection strategies.
Treatment and prevention
Optimal treatment for infectious/non-infectious subclinical TB and non-infectious clinical TB	Determining sufficiency of treatment for non-infectious TB.
Develop research networks that could undertake randomised trials in TB case finding	It is necessary to screen large numbers of individuals to identify new cases, so large multi-site approaches, with shared protocols, may potentially be more efficient.
Prevention of disease	Trials of vaccination to prevent disease should attempt to rule out subclinical TB at enrolment. The benefit/harm of vaccination is unknown if an individual has subclinical TB. Expanding vaccine trial outcomes to earlier states (i.e. beyond infectious clinical disease) may reduce time and costs of clinical studies.
Individual benefit from treating subclinical TB states
Impact on mortality and recurrence by subclinical TB treatment	Opportunities exist for the retrospective analysis of existing cohort data where people in subclinical TB states have received treatment.
Impact on post-TB sequelae by subclinical TB treatment	Evaluation of impact on post-TB sequalae requires multi-modal measurements embedded within subclinical TB treatment ttrials and follow up beyond the end of treatment.
Impact on psychosocial and economic quality of life by subclinical TB treatment	Engagement is needed to develop patient-informed outcomes that assess both desirable and undesirable outcomes.
Impact on comorbidity exacerbation and incidence by subclinical TB treatment	Timing and frequency of measurements needs to be defined by each clinically meaningful outcome. Opportunity for design of new trials for subclinical treatment to include collection of samples/data to assess these measures.
Population benefit from diagnosing and treating subclinical TB states
The potential and benefits of detection and treatment of subclinical TB to reduce transmission and incidence	Historical and contemporary case finding studies have suggested substantial impact on TB prevalence. But studies with clear disease phenotypes are key to provide direct empirical evidence.
Relative infectiousness of subclinical TB infectious states	Requires the identification of phenotypes in a defined population, with subsequent comparison of relative infectiousness metrics.
Tools to measure transmission	It is not possible to ‘prove’ transmission events, immunoreactivity is poorly understood, and new (blood-) based biomarkers of Mtb infection would be extremely welcome. These are likely to be challenging to develop.
Implementation Gaps
Programmatic implementation of sub-disease management	Implementation will require sufficient diagnostics, treatments and algorithms to be developed to avoid misclassification and inappropriate treatment
Translating new framework into practice	Requires significant changes to recording and reporting tools. Modification to guidelines, training and surveillance systems.
Cost-effectiveness of subclinical and non-infectious TB detection	To inform the likely economic benefits to TB program budgets, cost effectiveness evaluations will need to be conducted. The balance between resource-intensive mass screening initiatives and short-long term benefits to impact on clinical management costs needs to be weighed up.
Implications of population-based screening on individuals	Individuals and communities need to be engaged to evaluate potential stigma and implementation challenges, as well as benefits of integration in existing community initiatives and health screening programs.