Effect of isoniazid preventive therapy on risk of death in west African, HIV-infected adults with high CD4 cell counts: long-term follow-up of the Temprano ANRS 12136 trial

Background: Temprano ANRS 12136 was a factorial 2 × 2 trial that assessed the benefits of early antiretroviral therapy (ART; ie, in patients who had not reached the CD4 cell count threshold used to recommend starting ART, as per the WHO guidelines that were the standard during the study period) and 6-month isoniazid preventive therapy (IPT) in HIV-infected adults in Côte d'Ivoire. Early ART and IPT were shown to independently reduce the risk of severe morbidity at 30 months. Here, we present the efficacy of IPT in reducing mortality from the long-term follow-up of Temprano. Methods: For Temprano, participants were randomly assigned to four groups (deferred ART, deferred ART plus IPT, early ART, or early ART plus IPT). Participants who completed the trial follow-up were invited to participate in a post-trial phase. The primary post-trial phase endpoint was death, as analysed by the intention-to-treat principle. We used Cox proportional models to compare all-cause mortality between the IPT and no IPT strategies from inclusion in Temprano to the end of the follow-up period. Findings: Between March 18, 2008, and Jan 5, 2015, 2056 patients (mean baseline CD4 count 477 cells per μL) were followed up for 9404 patient-years (Temprano 4757; post-trial phase 4647). The median follow-up time was 4·9 years (IQR 3·3–5·8). 86 deaths were recorded (Temprano 47 deaths; post-trial phase 39 deaths), of which 34 were in patients randomly assigned IPT (6-year probability 4·1%, 95% CI 2·9–5·7) and 52 were in those randomly assigned no IPT (6·9%, 5·1–9·2). The hazard ratio of death in patients who had IPT compared with those who did not have IPT was 0·63 (95% CI, 0·41 to 0·97) after adjusting for the ART strategy (early vs deferred), and 0·61 (0·39–0·94) after adjustment for the ART strategy, baseline CD4 cell count, and other key characteristics. There was no evidence for statistical interaction between IPT and ART (pinteraction=0·77) or between IPT and time (pinteraction=0·94) on mortality. Interpretation: In Côte d'Ivoire, where the incidence of tuberculosis was last reported as 159 per 100 000 people, 6 months of IPT has a durable protective effect in reducing mortality in HIV-infected people, even in people with high CD4 cell counts and who have started ART. Funding: National Research Agency on AIDS and Viral Hepatitis (ANRS)