People living in high TB burden communities who undergo active case finding for TB may have chest X-ray changes suggestive of TB but their sputum test can be negative for the TB bacteria or they are unable to provide a sputum sample. This means they have bacteriologically unconfirmed TB. These individuals often have minimal or no symptoms and face uncertainty in clinical management-either receiving full 24-weeks TB treatment or often being discharged without treatment, risking loss to follow-up.

The RADIO-TB trial aims to address the following main questions in relation to people with bacteriologically unconfirmed TB.

1. What is the risk of progression if treatment is not immediately started but there is close and careful follow-up over 78 weeks?
2. If treatment is started immediately what is the shortest duration of TB treatment that is as effective as the standard 24-week TB treatment?
3. Are there better tests to confirm TB than the current routinely used, sputum-based tests?
4. What is the cost effectiveness of different treatment strategies and what might the impact on transmission be within communities?

Participants in the trial will be randomized into six groups: five receiving immediate standard TB treatment of varying durations (8 to 24 weeks), and one receiving deferred treatment with close monitoring. Deferred treatment will be initiated only if clinical deterioration occurs or TB is later confirmed. Participants will be closely followed for 78 weeks with at least 11 clinic visits.

Increasingly, countries are adopting community-based chest X-ray (CXR) screening strategies to identify people with TB at an early stage, often before they have symptoms. However, a significant proportion of individuals with radiographic abnormalities suggestive of TB are not bacteriologically confirmed due to limitations in sputum-based diagnostics or inability to produce sputum. These individuals are often diagnosed with "bacteriologically unconfirmed TB," and there is currently no consensus on how best to manage them.

The RADIO-TB trial is a multi-country, multi-arm, phase 3 clinical trial designed to address this critical evidence gap. The trial aims to determine the optimal treatment duration for individuals with radiographic evidence of TB who do not have bacteriological confirmation and are identified through active case finding (ACF) using CXR and computer-aided detection (CAD) software. The study is being conducted in South Africa, Zimbabwe, and Pakistan. Recruitment for the trial will be conducted alongside existing programatic CXR-based TB ACF activity taking place in these countries.

RADIO-TB uses a pragmatic, open-label, randomized design with six parallel arms. Participants are randomized to one of the following 6 arms in a ratio 1:1:1:1:1:2.

Arm A. Immediate start of 24 weeks (approx. 6 months) TB treatment comprised of 8 weeks (approx. 2 months) isoniazid, rifampicin, pyrazinamide and ethambutol (2HRZE) followed by 16 weeks (approx. 4 months) of isoniazid and rifampicin (4HR) - 2HRZE/4HR (control-A).

Arm B. Immediate start of 20 weeks (approx. 5 months) TB treatment - 2HRZE/3HR. Arm C. Immediate start of 16 weeks (approx. 4 months) TB treatment - 2HRZE/2HR. Arm D. Immediate start of 12 weeks (approx. 3 months) TB treatment - 2HRZE/1HR. Arm E. Immediate start of 8 weeks (approx. 2 months) TB treatment - 2HRZE. Arm F. No initial treatment, deferred if progression - if during close follow-up there is bacteriological confirmation and/or clinical/radiological progression, to start 24 weeks (approx. 6 months) TB treatment - 2HRZE/4HR (control-F).

The trial employs a Response Over Continuous Intervention (ROCI) design, which allows for the estimation and modelling of the duration-response curve and a more precise and efficient identification of the shortest non-inferior treatment duration compared to the standard 24-week regimen. The study hypothesis is that those with radiographically apparent, bacteriologically unconfirmed TB identified through ACF, immediate TB treatment with duration that is 8 weeks or longer, but less than 24 weeks, will be: (A) Non-inferior to treatment of 24 weeks duration (Arm A) at a 5% non-inferiority margin and (B) Superior to no initial treatment (Arm F) to prevent unfavourable TB outcomes.

The trial will enrol 784 participants aged 16 years and older who have undergone CXR-based TB ACF locally and found to have a CAD score above a locally defined threshold for TB investigation and either sputum negative by a WHO-approved rapid molecular test for TB or are unable to produce sputum. Key exclusion criteria are: a history of prior pulmonary TB, evidence of extra-pulmonary TB, clinical indication for immediate full-course treatment or an alternative cause for the CXR abnormalities more likely than TB. People who are HIV-infected will be eligible if their CD4 count is ≥200/mm3 and they on established on anti-retroviral therapy (ART). The CAD software used for screening will be those used locally in the existing ACF programmes according to international recommendation and national approvals. At the end of the study CXR will be re-analysed for all available CAD software to aid generalisability of findings.

All participants will undergo comprehensive baseline assessments, including clinical evaluation, induced sputum culture, blood and respiratory sampling for novel diagnostics. They are followed for 78 weeks with scheduled clinical evaluations, repeat chest X-ray, repeat microbiological testing, and quality-of-life assessments. Where necessary participants will be investigated for conditions other than TB.

The primary outcome measure for the trial is the proportion with unfavourable outcomes defined as requiring initiation of TB treatment after the week 0 visit for either bacteriologically confirmed TB or clinical/radiological progression without bacteriological confirmation. The trial also evaluates safety, tolerability of treatment; the impact on lung function, chest X-ray changes and quality of life; the frequency and timing of diagnoses other than TB; and the development of drug resistance. In addition, the study includes ancillary components to (1) assess the cost-effectiveness of different treatment strategies, (2) model the impact of different strategies on TB transmission and (3) evaluate novel diagnostic tools that may improve the early identification and confirmation of TB.

The RADIO-TB trial is sponsored by University College London (UCL) and funded by the Wellcome Trust. It is coordinated by the UCL Institute of Clinical Trials and Methodology (ICTM) in collaboration with national TB programs, academic institutions, and community partners in the participating countries. The trial incorporates robust patient and public involvement (PPI) mechanisms to ensure that the research is aligned with community priorities and that findings are disseminated in accessible and impactful ways.

By generating high-quality evidence on the optimal management of bacteriologically unconfirmed TB, the RADIO-TB trial aims to inform national and international TB treatment guidelines. The results will support more individualised and evidence-based approaches to TB care, improve patient outcomes, and contribute to global efforts to end the TB epidemic.

Inclusion Criteria:

1. Aged ≥16 years old on day of randomisation
2. ≥25kg in weight on day of randomisation
3. Participation in community or facility-based active case finding for TB by CXR
4. Digital CXR with Computer Aided Detection software score above threshold for sputum investigation for Mtb according to local guidance ≤21 days prior to randomisation
5. Attempted sputum collection for testing by WHO approved rapid molecular test for Mtb at active case finding and at screening

Exclusion Criteria:

1. Sputum sample from active case finding or screening confirmed as positive for Mtb by WHO approved rapid molecular test for Mtb
2. Previous history of pulmonary TB
3. TB preventive therapy taken within the last 6 months
4. >1 dose of treatment with anti-tuberculous activity taken in the 21 days prior to randomisation
5. HIV infected, unless CD4 ≥200 cells/mm3 and on anti-retroviral therapy for at least 8 weeks
6. Pregnancy (confirmed by high sensitivity urine test)
7. Contraindication to study drug
8. Investigator concern for potential extra-pulmonary TB with the exceptions of (a) intrathoracic lymphadenopathy; and/or (b) small pleural effusion of limited extent, neither felt to require invasive diagnostic sampling
9. CXR changes considered more likely to relate to a condition other than untreated TB by investigator
10. Investigator assessment of an immediate requirement for full 24-weeks standard TB treatment
11. Investigator assessment of unacceptable risks to potential participant from study entry
12. No fixed home address
13. Unable to participate in the full duration of study follow-up

Plan to upload IPD to public repository e.g. WHO IPD or TB-PACTS

After primary publication (est Q4 2029)

Via application to the IPD repository