Background: Once-weekly isoniazid and rifapentine for 3 months (3HP) by directly observed therapy (DOT) is safe and effective for treating latent tuberculosis infection (LTBI). Treatment completion with 3HP by DOT was 82.1% in the TBTC Prevent TB study. Implementation of 3HP is limited by the requirement for DOT. Models show that treatment completion by self-administered therapy (SAT) could be lower and still be cost-effective. The iAdhere study compared 3HP completion rates by DOT versus SAT or enhanced SAT with weekly text reminders (eSAT).
Methods: The study was an international clinical trial among adults with LTBI and no contraindications for 3HP or SAT. Randomization was 1:1:1, stratified by site. A non-inferiority margin of 15% was used based on cost-effectiveness modeling in the US, and enrollment targeted >75% from the US to have power for a pre-planned sub-analysis. The primary outcome was completion of > 11 doses within 16 weeks, as determined by clinic dose records and pill counts for DOT, and by self-reports, pill counts, and medication event monitoring system (MEMS) data for SAT and eSAT.
Results: Of 1,002 patients enrolled, 4 were excluded as contacts to drug-resistant TB, 998 were eligible to complete treatment, and 772 (77%) were enrolled in the US. The study arms were demographically similar. Median age was 37 years [IQR 27, 49]. Participants included 482 (48%) women, 344 (34%) contacts to active TB, and 141 (14%) LTBI test converters; 85 (8%) had diabetes, 11 (1%) were HIV positive, 776 (78%) HIV negative, and 215 (21%) HIV unknown. Overall treatment completion was 87.2% [95%CI 83.1%-90.5%] by DOT, 74.0% [68.9%-78.6%] by SAT, and 76.4% [71.3%-80.8%] by eSAT. Treatment completion in US participants was 85.4% [80.4%-89.4%], 77.9% [77.2%-82.6%], and 76.7% [70.9%-81.7%] respectively. SAT was non-inferior to DOT in the US but not overall and eSAT did not achieve non-inferiority (figure). Discontinuation rates due to adverse effects were similar by arm, 3.6% DOT, 5.4% SAT, 4.3% eSAT (P=0.52).
Conclusions: Our findings support the use of 3HP by SAT in the US. Non-inferiority was not established for SAT or eSAT overall due to higher than predicted DOT completion rates and variability in SAT and eSAT completion outside the US. Further cost-effectiveness analyses and evaluation of the role of text reminders are needed.