The optimal time to initiate antiretroviral therapy (ART) in HIV-associated tuberculous meningitis (TBM) unknown. There are concerns that immediate ART may worsen rather than improve outcome, because drug interactiond and toxicities or development of an intracerebral immune reconstitution inflammatory syndrome (IRIS). Conversely, delaying ART may result in increased HIV-related deaths. To answer this question, we are conducting a randomised, double-blind placebo-controlled trial comparing immediate and deferred ART in HIV-infected patients presenting with TBM, to assess effect on survival.

Title: Study of immediate versus deferred antiretroviral therapy in HIV-associated tuberculous meningitis Study design: A randomized, double blind, placebo-controlled trial with 2 parallel arms Sample size: 247 Inclusion criteria: Age 15 years or older; HIV antibody positive; clinical diagnosis of TBM.

Exclusion criteria: positive CSF Gram or India ink stain, known or suspected pregnancy; antituberculous treatment 8 to 30 days immediately prior to recruitment; previous antiretroviral therapy; laboratory contraindications to antiretroviral or antituberculous therapy; lack of consent.

Consent: Written informed consent will be sought for all patients. Verbal consent will be considered acceptable when written consent is impossible. In unconscious patients, the consent of 2 independent physicians will be considered acceptable.

Randomization: Patients will be stratified according to TBM disease severity at presentation (modified MRC grade I to III). Within each stratum, patients will be randomized to 1 of the 2 treatment arms: immediate or deferred (2 months) ART.

Antituberculous treatment: Initial therapy will be with isonazid, rifampicin, pyrazinamide and ethambutol for 3 months. After 3 months, patients will continue on rifampicin and isoniazid for a further 6 months.

Corticosteroid treatment: Dexamethasone 0.3 - 0.4mg/kg will be administered and tapered over 6 - 8 weeks, according to TBM grade.

Antiretrovira l treatment: Antiretrovirals (zidovudine, lamivudine and efavirenz)or identical placebo tablets will be commenced at study entry and continued for 2 months. Thereafter, all patients will received antiretrovirals.

Clinical monitoring: Patients will be assessed weekly as an inpatient for 3 months. Hospital outpatient review will occur monthly until 9 months. A final follow-up visit will take place at 12 months.

Laboratory monitoring: Routine laboratory tests will be monitored weekly as an inpatient and monthly as an outpatient. Blood samples for CD4 T-lymphocyte count and plasma HIV-1 RNA level will be monitored 3-monthly. CSF samples will be taken at 0, 1, 2, 3, 6 and 9 months.

Radiology: Patients will have a chest radiograph performed on admission. A CT or MRI brain scan may also be performed if clinically indicated.

Adverse events: All grade 3 and 4 adverse events will be reported immediately to the Data and Safety Monitoring Committee.

Outcome measures: The primary endpoint will be mortality at 9 months. The secondary endpoints will be: mortality at 12 months; fever clearance time; coma clearance time; neurological relapse; progression to new or recurrent AIDS defining illness; any grade 3 or 4 adverse event; CD4 count response; plasma HIV-1 RNA response; neurological disability.

Data analysis: Analysis will be based on intention to treat.

Inclusion Criteria:

• age 15 years or older
• HIV antibody positive
• clinical diagnosis of TB meningitis

Exclusion Criteria:

• positive CSF Gram or India ink stain
• known or suspected pregnancy
• antituberculous treatment 8 - 30 days immediately prior to recruitment
• previous antiretroviral therapy
• laboratory contraindications to antiretroviral or antituberculous therapy
• lack of consent.