Giving treatment for tuberculosis to all people with very advanced HIV disease at the time they start antiretroviral therapy, before a laboratory diagnosis of TB – so-called empiric treatment – does not reduce the risk of death compared to giving antiretroviral therapy and isoniazid preventive therapy. These findings were presented by Dr Mina Hosseinipour at the 8th International AIDS Society Conference on HIV Pathogenesis, Treatment and Prevention (IAS 2015).
Giving TB therapy in advance of a laboratory-confirmed diagnosis, in populations with a high burden of tuberculosis, is called empiric therapy. Physicians might choose to give empiric treatment where patients are very sick, where they suspect TB and when immediate initiation of treatment may be necessary. The World Health Organisation recommends empiric therapy in cases of suspected extrapulmonary tuberculosis.
But the case for empiric TB treatment extends beyond treating those people who turn up with very obvious symptoms of tuberculosis, in the view of some experts. A systematic review published in 2011 found that a median of 17% of people starting antiretroviral therapy developed TB within one year, and those starting treatment at CD4 counts below 50 cells/mm3 were at especially high risk.
Providing empiric TB treatment to everyone starting antiretroviral therapy could place an additional burden on health systems where drug supply systems are overstretched, but empiric treatment for people at the highest risk of dying of TB might be more feasible. But is it more effective than giving isoniazid preventive treatment alongside antiretroviral therapy? This was the question investigated in the REMEMBER (Reducing Early Mortality and Early Morbidity by Empiric Tuberculosis Treatment Regimens) trial (ACTG 5274).
REMEMBER was conducted in 11 countries with a high burden of TB in sub-Saharan Africa, Asia, the Caribbean and Latin America. The study recruited people with HIV with CD4 cell counts below 50 cells/mm3 about to initiate antiretroviral therapy and randomised then to receive either empiric TB therapy (a standard 4-drug regimen) or isoniazid preventive therapy for 24 weeks, with TB therapy given if active TB was diagnosed.
The study excluded people with active or suspected TB, so it was a test of the effect of empiric treatment on the prevention of TB in people who might have very early active disease that was not yet presenting symptoms, and the prevention of the development of active TB in the first year after starting treatment, when immune reconstitution is still limited. The study also excluded people who had received any treatment for active TB in the preceding 96 weeks, and anyone who had received more than 30 days of isoniazid preventive treatment in the previous 48 weeks. The primary study endpoint was survival at 24 weeks after randomisation; secondary analyses compared TB incidence, time to death and safety and tolerability.
The study enrolled 850 people randomised equally to the two study arms. Participants were predominantly Black (90%) and had very advanced HIV disease (median CD4 cell count 18 cells/mm3). Just over 60% of all participants had a CD4 cell count below 25 cells/mm3. The study found a lower-than-expected rate of TB, and no significant difference between the two study arms in the risk of death after 24 weeks (4.8% in the empiric arm vs 5.2% in the IPT arm, a non-significant difference).
Participants were stratified by CD4 count above or below 25, by enrolment before or after the introduction of Xpert MTB/RIF testing of sputum samples at baseline, and by the presence of poor prognostic factors. None of these factors significantly influenced the risk of death.
An analysis of the proportions developing TB by week 24 found a significantly higher frequency in the empiric treatment group (p = 0.001). TB was diagnosed more frequently in the empiric treatment arm (33 cases vs 19), with the difference largely accounted for confirmed cases of pulmonary tuberculosis. No difference was observed in adverse events, including TB-related immune reconstitution inflammatory syndrome (IRIS).
The investigators concluded that the current WHO recommendation of antiretroviral therapy plus IPT is sufficient to reduce the risk of developing active TB in people with very advanced HIV disease, and that where systematic screening for TB is taking place in line with WHO guidance, empiric TB treatment is not necessary.