Isoniazid preventive therapy for HIV-infected people: evidence to support implementation

Tuberculosis (TB) is a common and deadly disease in people living with HIV worldwide. In most HIV prevalent countries, HIV is the predominant driver of the TB epidemic. Decades of progress in TB control have been reversed or slowed by failure to identify, prevent and treat TB in HIV-infected persons, their family and communities. Although successful antiretroviral therapy (ART) reduces TB risk, opportunities for detection and prevention of TB in HIV care settings are often missed. In countries with high TB and HIV burdens, the World Health Organization recommends intensified TB case finding, isoniazid preventive therapy (IPT) and infection control for TB, which are branded as the three I's for HIV/TB [1,2]. Comstock et al. [3] conducted the first trials of community-wide IPT in Bethel, Alaska, in the 1950s and 1960s, demonstrating a 70% reduction in TB incidence with widespread provision of isoniazid. Recommendations for use of IPT for persons with HIV infection were first made by International Union Against Tuberculosis and Lung Disease (IUATLD) and WHO in 1993 [4], revised in 1998 [5] and further strengthened in 2004 [2]. The WHO issued new evidence-based guidelines that reconceptualized IPT and combined it with TB screening in 2010 [6]. Yet in the 2009 WHO TB Control Report, the WHO estimates that fewer than 0.5% of HIV-infected persons worldwide have received IPT [7]. Why is uptake of IPT so poor? There are both perceived barriers and real operational challenges. Anecdotally, clinicians are concerned that failure to rule out active disease will contribute to isoniazid resistance, and ultimately lead to multidrug resistance. Clinicians are also concerned about toxicity of isoniazid and maintaining adherence to the IPT regimen. From a programmatic standpoint, it is often not clear if the HIV program or the TB control program will take responsibility for IPT. Standardized screening algorithms for ruling out TB on the basis of sound scientific data have not been available until recently [6], and suboptimal diagnostic tools raise concerns about missing TB disease. This supplement to AIDS addresses many of the concerns about implementing IPT, including safety, resistance and feasibility, mostly within the community-level intervention projects of the Consortium to Respond Effectively to the AIDS-TB Epidemic (CREATE). Established in 2004 and funded by the Bill and Melinda Gates Foundation, CREATE is conducting cluster-randomized trials of TB/HIV interventions in Brazil, South Africa and Zambia. Unique to research consortia, CREATE has incorporated national and global policy and advocacy activities since its inception; partnering with WHO to promote TB/HIV research, educate media, engage HIV and TB researchers and advocates and influence national and global policy. Getahun et al. [8] describe the global progress, challenges in scaling-up and the factors essential for success in IPT implementation. Each study is unique in its design and targeted population; however, all share the goal of utilizing existing public health tools to reduce the morbidity and mortality from HIV-related TB. All studies also conduct intensified case finding, a necessary prerequisite for implementing IPT. The largest CREATE IPT implementation trial, Thibela TB, is a cluster-randomized trial of community-wide IPT among gold miners in Gauteng, Free State and North West provinces in South Africa (Fielding KL. Manuscript submitted). The goal of the study is to reduce TB incidence in the participating mine shafts by identifying and treating active TB, reducing reactivation of latent TB and prevention of early progression of newly acquired infection. Gold miners have one of the highest TB rates in the world, exacerbated by silicosis, HIV co-infection and close living and working conditions. In a pilot study of gold mines participating in the Thibela TB trial, an estimated 89% miners had latent TB infection [9]. Previous studies among South African gold miners have reported that nearly one third of gold miners have HIV infection [10]. In Thibela TB, mine shafts are randomized to receive either the mass intervention, consisting of community mobilization and education, TB screening and IPT for all individuals without a contraindication, or the standard of care, which includes at least annual screening using chest radiograph and IPT for high-risk miners (HIV-infected and silicotics). Over 27 000 miners have been screened for TB, and more than 24 000 have started IPT over 36 months. Trial results are due in 2012. In the process of implementing IPT, the Thibela TB trial identified more cases of active tuberculosis than the existing healthcare system; this highlights the benefit of conducting intensified TB screening before initiating IPT [11]. Articles in this supplement describe experiences of Thibela TB including factors associated with undiagnosed TB [12], adverse events associated with IPT [13] and community mobilization strategies to support implementation of community-wide IPT [14]. The Thibela TB investigators report also on a qualitative study [15] of perceived barriers to IPT among South African physicians; a study of the combined effect of IPT with ART on early mortality in a large workplace ART program [16] and a commentary on how an effect of an IPT program on mortality may have been underestimated in randomized controlled trials [17]. The THRio study in Rio de Janeiro, Brazil, is implementing and evaluating tuberculin skin testing and IPT in 29 municipal HIV clinics. Brazil offers free ART through its public HIV clinics. National guidelines have long recommended the use of tuberculin skin test for TB screening and use of IPT for preventing TB in HIV-infected people, but these guidelines have largely not been operationalized. The THRio intervention involved education of physicians, nurses and staff on TB prevention guidelines using a step-wedge implementation design and comparing TB rates among patients in clinics that have received the intervention to controls that have not yet received it [18]. The intervention is now complete with all 29 clinics having received the intervention. Over 5000 HIV-infected patients have had tuberculin skin tests placed; 94% were read and 18% were positive. Eighty-three percent of those with positive skin tests have completed therapy to date [19]. Although the primary outcome data of TB incidence is pending, a baseline study conducted in these clinics revealed that combined ART and IPT is more effective in reducing TB incidence than either drug used individually [20]. In this supplement, Durovni et al. [19] present data on the use of tuberculin skin test and the utility of secondary IPT prophylaxis. The third study of the CREATE consortium is the Zambia and South Africa TB AIDS Reduction Study, ZAMSTAR. This community-randomized trial conducts innovative intensified case finding across 24 communities using social mobilization, increased access to TB diagnostics and integrated TB/HIV care for households of people living with active TB [21]. In collaboration with the Zambian National TB program, IPT pilots are underway in ZAMSTAR clinic and household sites. This supplement provides ample operational data on feasibility, adherence, safety and resistance to inform IPT policy and practice and promote large-scale roll out of IPT. Each of these studies individually has the potential to inform clinical care, strengthen public health practice and contribute to TB/HIV policy. As a consortium, CREATE has modeled community engagement and mobilization, strengthened existing healthcare infrastructures and advocated for stronger national and global TB/HIV policies, while conducting innovative community-wide intervention research. For CREATE, the sum is greater than its parts. Acknowledgement This study is funded by the Bill and Melinda Gates Foundation. Conflicts of interest: None.