Treatment of latent tuberculosis (TB) infection (LTBI) effectively prevents its progression to active TB. However, long treatment duration and drug-related hepatotoxicity limit the effectiveness of the 9-month daily isoniazid (9H). Data on the 3-month weekly rifapentine plus isoniazid (3 HP) in Asian populations are currently unavailable. We prospectively randomised the LTBI contacts aged ≥12 years with positive tuberculin skin test into 9H and 3 HP groups in four hospitals between January 2014 and May 2016 in Taiwan. The primary and secondary outcomes were treatment completion rate and adverse drug reactions (ADRs), respectively. Overall, 263 participants with LTBI were randomised into the 3 HP (n = 132) and 9H groups (n = 131); 14 (10.6%) and 29 (22.1%) participants in the 3 HP and 9H groups, respectively, discontinued therapy (p = 0.011). Discontinuation rates owing to ADRs were 9.1% (3 HP) and 5.3% (9H) (p = 0.241). Clinically relevant hepatotoxicity was more common in the 9H than in the 3 HP group (5.3% vs. 1.5%; p = 0.103), whereas systemic drug reaction was more common in the 3 HP than in the 9H group (3.8% vs. 0%; p = 0.060). Women had a significantly higher rate of Grade II fever than men (13.7% vs. 1.2%; p = 0.003). Compared with the 9H regimen, the 3 HP regimen had a higher completion rate with lower hepatotoxicity and well-tolerated ADR.

number NCT02208427.