Abacavir, zidovudine, or stavudine as paediatric tablets for African HIV-infected children (CHAPAS-3): an open-label, parallel-group, randomised controlled trial

BACKGROUND WHO 2013 guidelines recommend universal treatment for HIV-infected children younger than 5 years. No paediatric trials have compared nucleoside reverse-transcriptase inhibitors (NRTIs) in first-line antiretroviral therapy (ART) in Africa, where most HIV-infected children live. We aimed to compare stavudine, zidovudine, or abacavir as dual or triple fixed-dose-combination paediatric tablets with lamivudine and nevirapine or efavirenz. METHODS In this open-label, parallel-group, randomised trial (CHAPAS-3), we enrolled children from one centre in Zambia and three in Uganda who were previously untreated (ART naive) or on stavudine for more than 2 years with viral load less than 50 copies per mL (ART experienced). Computer-generated randomisation tables were incorporated securely within the database. The primary endpoint was grade 2-4 clinical or grade 3/4 laboratory adverse events. Analysis was intention to treat. This trial is registered with the ISRCTN Registry number, 69078957. FINDINGS Between Nov 8, 2010, and Dec 28, 2011, 480 children were randomised: 156 to stavudine, 159 to zidovudine, and 165 to abacavir. After two were excluded due to randomisation error, 156 children were analysed in the stavudine group, 158 in the zidovudine group, and 164 in the abacavir group, and followed for median 2·3 years (5% lost to follow-up). 365 (76%) were ART naive (median age 2·6 years vs 6·2 years in ART experienced). 917 grade 2-4 clinical or grade 3/4 laboratory adverse events (835 clinical [634 grade 2]; 40 laboratory) occurred in 104 (67%) children on stavudine, 103 (65%) on zidovudine, and 105 (64%), on abacavir (p=0·63; zidovudine vs stavudine: hazard ratio [HR] 0·99 [95% CI 0·75-1·29]; abacavir vs stavudine: HR 0·88 [0·67-1·15]). At 48 weeks, 98 (85%), 81 (80%) and 95 (81%) ART-naive children in the stavudine, zidovudine, and abacavir groups, respectively, had viral load less than 400 copies per mL (p=0·58); most ART-experienced children maintained suppression (p=1·00). INTERPRETATION All NRTIs had low toxicity and good clinical, immunological, and virological responses. Clinical and subclinical lipodystrophy was not noted in those younger than 5 years and anaemia was no more frequent with zidovudine than with the other drugs. Absence of hypersensitivity reactions, superior resistance profile and once-daily dosing favours abacavir for African children, supporting WHO 2013 guidelines. FUNDING European Developing Countries Clinical Trials Partnership.