• TB incidence
  
• mortality
  

• intercurrent infections
  
• adherence
  
• adverse events
  
• antimicrobial resistance
  

Tuberculosis (TB) and HIV are dual pandemics occurring in South Africa. Prevention of TB and the subsequent decline in immune function in HIV-infected children is an important strategy to reduce mortality. Isoniazid (INH) prophylaxis reduces TB incidence in HIV-infected adults, but the efficacy in HIV-infected children has not been studied. In 2003, we therefore began a double blind placebo controlled trial to investigate the impact of INH prophylaxis on mortality, morbidity and TB incidence in HIV-infected children. Interim analysis found a striking reduction in mortality and TB with a decrease in mortality in excess of 50% and 60% respectively, in children on INH. Based on this, the placebo arm was terminated; the study continued as a trial of thrice versus daily INH and cotrimoxazole (CTX). Although the results indicate an important benefit in children on INH, it is unknown what the long term efficacy and safety of INH prophylaxis is, what the optimal regime is and whether this pertains to children on HAART (who formed a minority of the cohort but who are still at risk for TB).

Aim To investigate the efficacy, safety and tolerability of INH and CTX as prophylactic strategies for HIV-infected children in a high TB prevalence area.

Objectives

1. To compare the long term impact of two different INH preventive regimens (daily versus thrice weekly) on TB incidence, occurrence of INH resistance in patients with culture-confirmed TB, mortality, incidence of adverse reactions and adherence
2. To compare the long term impact of two different CTX prophylactic regimens (daily versus thrice weekly) on mortality, frequency and duration of hospitalization, type of serious infections, nasopharyngeal carriage of bacteria and development of antimicrobial resistance, adherence and incidence of adverse reactions
3. To investigate the efficacy, safety and tolerability of INH compared with placebo for prevention of TB in children receiving HAART

Methods A prospective randomized double blind placebo controlled study of INH versus placebo in newly recruited HIV-infected children who are stable on HAART. In addition, an extended follow-up study of children already randomised to thrice weekly or daily INH and CTX. Children will be followed for 2 years with regular clinical evaluation, adherence assessment and laboratory monitoring. Outcomes measured will be mortality, TB incidence, morbidity, adherence and tolerability.

Inclusion Criteria:

• HIV-infected children
• Resident in Cape Town
• Informed consent obtainable
• weight > 2.5kg
• Access to transport
• HAART use for not less than 2 months but not more than 12 months with no significant demonstrated toxicity and good adherence

Exclusion Criteria:

• Chronic diarrhoea
• Current use of INH prophylaxis
• Prior hypersensitivity to INH prior history of allergy to sulphur drugs
• Prior history of allergy to sulphur drugs
• Severe anaemia (haemoglobin less than 7 gm/dl)
• Neutropenia (absoloute neutrophil count less than 400 cells)
• Thrombocytopenia (platelet count < 50 000/uL)
• Non-reversible renal failure
• Clinical hepatitis
• Exposure to household TB contact, requiring INH prophylaxis