This study is a parallel, two part, open label, individually randomized, pragmatic trial among HIV-positive individuals. Part A compares a single round of weekly high dose rifapentine plus isoniazid for three months (3HP) to six months of daily isoniazid (6H). Part B compares periodic 3HP (p3HP) to a single round of 3HP.

Part A: A randomised controlled trial of 3HP vs 6H [enrollment starts concurrently with Part B]

Justification: The World Health Organization (WHO) recommends at least six months of isoniazid (6H) for persons living with HIV. However, 6H remains poorly implemented in most high burden tuberculosis (TB) countries. In its 2015 guidelines, WHO includes 3HP as an latent tuberculosis infection (LTBI) treatment option for high-income and upper middle-income countries with TB incidence rates <100/100,000. One trial comparing 3HP to 6H in a high burden country suggests that a single round of 3HP has less toxicity, better treatment completion rates, and similar efficacy in preventing TB. The purpose of comparing a single round of 3HP to 6H is to demonstrate the feasibility of implementing 3HP in high burden countries, to explore its safety and effectiveness, and to generate evidence to guide a WHO recommendation for the use of 3HP in high burden settings.

Sample size: If we assume 85% of patients in the 6H arm complete treatment as defined above, with 400 patients in the 6H arm and 3600 patients in the 3HP arm we will have 82% power to detect an increase in treatment completion of 5% (To -90%) in the 3HP arm. If treatment completion in the 6H arm is 75%, we will have approximately 90% power to detect an increase in treatment completion of 7% in the 3HP arm.

Analysis: Treatment completion will be compared by study arm using Fishers Exact test, and associated risk difference and 95% confidence interval (CI).

Part B: A randomised controlled trial of 3HP vs p3HP [enrollment starts concurrently with Part A]

Justification: A single round of 3HP has been shown to be non-inferior to 9 months of isoniazid (9H) in persons at high risk of developing TB in low and middle TB burden settings. Similarly, a single round of 3HP has demonstrated similar efficacy in preventing active TB when compared to 6H among HIV-positive, tuberculin skin test (TST)-positive adults in the high burden setting of South Africa. In high burden settings, 6H and 3HP provide protection of limited duration probably due to high ongoing transmission and reinfection. Continuous isoniazid preventive therapy has been shown to provide more durable protection in high burden settings, but is not currently policy outside of a handful of countries, and the actual uptake is poor. Giving 3HP periodically may provide durable protection, be easier for health systems to implement, and may be associated with better adherence and fewer side effects.

Sample size: Assuming a cumulative TB incidence of 5% over 2 years in the control (3HP) arm, an overall loss to follow-up of 10% by year 2, a two-sided type I error of 5%, 1:1 randomisation, a superiority comparison and 1800 participants per arm, we have 80% power to detect a 40% reduction in cumulative TB incidence from months 0 to 24.

Analysis: The analysis will compare 3HP and p3HP with two years of follow up. Cumulative TB incidence will be determined by combining incident TB cases identified over the 24 months of follow up AND prevalent TB cases identified at the 12 and 24 month culture survey. Data will be reported as a risk difference and odds ratio and their associated 95% CIs, adjusting for randomisation strata. The secondary outcome comparing the effectiveness of p3HP to 3HP from month 13 to 24 (during which time the greatest effect is likely to be evident) will be conducting using the same analytic methods. The results of Part B will be disseminated subsequent to the results of Part A.

Inclusion Criteria:

• At least two years of age
• Known HIV infection
• Antiretroviral therapy (ART) ineligible or on ART for ≥3 months

Exclusion Criteria:

• Confirmed or suspected TB disease
• Likely to move from the study area during the study period
• Known exposure to TB cases with known or suspected resistance to isoniazid or rifampicin in the source case
• TB treatment within the past year
• TB preventive therapy within the last year
• Sensitivity or intolerance to isoniazid or rifamycins
• Suspected acute hepatitis or known chronic liver disease
• ALT/AST >5 times the upper limit of normal (regardless of symptoms of hepatitis)
• Pregnancy or breastfeeding
• Women of childbearing potential who are unable or unwilling to use contraception
• Self-reported alcohol use exceeding 28 units per week for men, or 21 units for women