In countries with a high tuberculosis (TB) prevalence, TB and invasive bacterial infections are leading causes of early death in patients who initiate antiretroviral therapy (ART) with advanced immunodeficiency.

We hypothesize that a systematic 6-month empirical TB treatment initiated 2 weeks before the introduction of ART in HIV-infected adults with severe immunosuppression (CD4<100/mm3) and no overt evidence of TB will reduce the risk of death and invasive bacterial infections. This strategy will be compared to one of extensive TB testing using point-of-care tests (Xpert MTB/RIF® and urine lipoarabinomanan LAM) and chest X-ray to identify and treat only patients with at least one positive test suggestive of TB.

Settings: Cambodia, Côte d'Ivoire, Uganda, Vietnam. Design: Multicentre, two-arm, unblinded randomized controlled superiority trial.

Objective: To compare the 24-week risk of death and occurrence of invasive bacterial infection between two experimental strategies in HIV-1 infected adults who start ART with a CD4 count <100/mm3: (i) continuous extensive TB screening during follow-up each time the patient present with symptoms, versus (ii) systematic empirical TB treatment started 2 weeks before ART initiation.

Trial strategies:

At inclusion, participants will be randomized 1:1 in two strategies of TB testing and treatment: extensive TB screening, or systematic empirical TB treatment.

Extensive TB screening (arm 1): In this arm:

• TB screening point-of-care tests (Xpert MTB/RIF®, urine LAM) and chest X-ray will be used extensively at randomisation (in all patients) and during follow-up (in patients with signs or symptoms suggestive of TB);
• Only patients who meet standardized criteria for TB at inclusion or during follow-up will receive a standard TB treatment (2ERHZ/4RH);
• ART (tenofovir(TDF)-lamivudine (3TC)/emtricitabine(FTC) or zidovudine (AZT)-lamivudine+ efavirenz) will be started immediately after randomization in patients not put on TB treatment, and 2 weeks after initiation of TB treatment in others.

Systematic empirical TB treatment (arm 2): In this arm:

• TB screening point-of-care tests will not be used;
• All patients will start a 6-month standard TB treatment (2ERHZ/4RH) at randomization; ART (tenofovir-lamivudine/emtricitabine or zidovudine-lamivudine+ efavirenz) will be started 2 weeks after TB treatment initiation.

Both strategies will apply to the first 24 weeks in the trial (intervention period).

From week-24 to week-48, the choice of TB tests and the prescription of TB treatment will be left upon the decision of the investigator in both trial arms.

Inclusion time: 24 months. Follow-up: each patient will be followed 48 weeks. Statistical analysis: the primary analysis will be intention to treat. It will compare the 24-week probability of death or invasive bacterial infection between arms.

Sample size: 1050 participants. This will allow demonstration of a 40% reduction in the 24-week probability of death or invasive bacterial infection in arm 2, compared to arm 1 (α 5%; 1-β 80%).

Inclusion Criteria:

• Age ≥18 years;
• HIV-1 infection as documented at any time prior to trial entry, as per national testing procedures;
• CD4 <100 cells/mm3;
• No history of antiretroviral drug use (except transient ART for PMTCT);
• Able to correctly understand the trial and to sign the informed consent.

Exclusion Criteria:

• HIV-2 co-infection;
• Contra-indication to efavirenz;
• Aspartate aminotransferase (AST) or Alanine aminotransferase (ALT) >5 times the upper limit of normal;
• Creatinine clearance <50 ml/min;
• Overt evidence that TB treatment should be started immediately;
• History of TB treatment in the past 5 years;
• Ongoing TB chemoprophylaxis (isoniazid preventive therapy);
• Any condition that would lead to differ ART initiation (e.g. acute condition requiring investigations and/or treatment prior to ART initiation);
• Current pregnancy or breastfeeding.