Background:  No trials have reported on the efficacy of cotrimoxazole prophylaxis for adults or children with HIV in areas of high levels of bacterial resistance to the drug.

Methods:  Children with HIV Antibiotic Prophylaxis (CHAP) is a randomized trial comparing daily cotrimoxazole (240 mg until 5 years, then 480 mg) with matching placebo in HIV-infected children aged 1 to 14 years in Zambia. Primary outcomes were mortality and grade 3 or 4 adverse drug reactions (ADR) thought possibly related to blinded drug. Hospital admissions were a secondary outcome. All efficacy comparisons were as randomized (intent-to-treat).

Results:  In October 2003, the trial was stopped early, after 541 children had been randomized (268 cotrimoxazole [C], 273 placebo [P]); 7 (3 C, 4 P) were subsequently determined to be HIV negative and excluded. Vital status after randomization was ascertained on all but 1 child (P), and 89.0% (C) and 86.3% (P) of child-time at risk was spent on blinded drug in the two arms respectively. At randomization, median age was 4.4 (IQR 2.1 to 8.3) years with 82 (15%) over 10 years; 50% were boys; median CD4% was 11% (n = 393, IQR 6 to 16%). During a median of 19 (IQR 14-24) months follow-up, 74 (28%) children died in the cotrimoxazole arm and 112 (42%) in the placebo arm (HR = 0.57 [95%CI:  0.43 to 0.77] p = 0.0002). This benefit persisted from 12 months onwards (n = 320, HR = 0.48 [0.27 to 0.84] p = 0.009), was present across all age groups (p = 0.82 for heterogeneity) and similar among children with baseline CD4% <15% (HR = 0.51 [0.34 to 0.77]) and >15% (HR = 0.62 [0.26 to 1.47], heterogeneity p = 0.73). In terms of ADR, there was 1 rash (P), 23 anemias (10 C, 13 P), and 12 neutropenias (6 C, 6 P) (logrank p = 0.43). 89 (34%) vs 117 (43%) children in the cotrimoxazole and placebo arms had at least 1 hospital admission (p = 0.02); total inpatient days were 1147 (C) and 1729 (P). PCP was diagnosed in only one child (P) and 73 nasopharyngeal aspirates were negative for Pneumocystis Carinii by PCR.

Conclusions: Our results suggest that HIV-infected children of all ages and with all levels of CD4% should receive cotrimoxazole prophylaxis. Economic analyses are planned to evaluate this approach, which could be carried out at community level. As countries prepare to provide antiretroviral drugs, CHAP trial results should provide impetus for HIV testing and for provision of cotrimoxazole and nutritional support for all HIV-infected children, irrespective of local levels of cotrimoxazole resistance.

Keywords: Cotrimoxazole; Mortality; Children