Background:  In 2001, Botswana implemented a national isoniazid preventive therapy program (IPT) for tuberculosis (TB) among HIV-infected adults (PLWH); each client receives 6 months’ IPT regardless of tuberculin skin test (TST) status. PLWH in TB-endemic countries suffer high rates of re-infection and TB disease and may benefit from continuous IPT.

Methods:  During 2004 to 2006, we recruited 1995 PLWH in Gaborone and Francistown from government facilities where IPT is provided. After a 6-month course of open-label IPT, participants received either isoniazid or placebo for an additional 30 months in a randomized, double-blind fashion. Per IPT Program guidelines, exclusion criteria included:  current illness (e.g., fever, cough), lymphadenopathy, jaundice or pulmonary disease on physical exam, history of AIDS-defining illnesses or active TB in the previous 3 years, and prior IPT. Additional trial-related requirements included:  a chest radiograph, routine blood, liver and kidney tests, and a negative pregnancy test. A minimum CD4 lymphocyte count was not required. Participants had access to free ART. A modified intent-to-treat analysis was used to compare the incidence of TB and adverse events in the 2-arm study.

Results:  At enrolment, participants’ median age was 32 years, 72% were female, 24% were TST⁺ and median CD4 count was 294 cells/mm³. During 5325 person-years of observation, participants attended a median of 35 of 36 maximum visits and 15 participants were lost to follow-up. ART was initiated for 52% of participants. Regarding outcomes and toxicity there were: 64 incident TB cases (52% confirmed by culture); 58 severe adverse events (67% hepatitis, 12% rash) with 59% occurring during the first 6 months; and 69 deaths—38% were AIDS-related and one was probably associated with IPT.

Conclusions:  The trial concluded successfully with high retention and sufficient endpoints. Efficacy and safety analyses are pending the unblinding of codes.