Résumé
Objectives:Perinatal single-dose nevirapine (sdNVP) selects for resistance mutations. The objective of this trial was to compare two maternal sdNVP-sparing regimens with standard zidovudine (ZDV)/sdNVP prophylaxis.Design:PHPT-5 was a randomized, partially double-blind placebo-controlled, noninferiority trial in Thailand (NCT00409591). Study participants were women with CD4(+) of at least 250cells/l and their infants.Methods:All women received ZDV from 28 weeks' gestation and their newborn infants for one week. Women were also randomized to receive NVP-NVP (reference): maternal intrapartum sdNVP with a 7-day tail' of ZDV along with lamivudine, and infant NVP (one dose immediately, another 48h later); infant-only NVP: maternal placebos for sdNVP and the tail', with infant NVP; LPV/r: maternal LPV/r starting at 28 weeks. Infants were formula-fed. HIV-diagnosis was determined by DNA-PCR.Results:Four-hundred and thirty-five women were randomized between January 2009 and September 2010. Accrual was terminated prematurely following a change in Thai guidelines recommending antiretroviral combination therapy for all pregnant women. Data on 405 mothers and 407 live-born children were analyzed. Baseline characteristics were similar between arms. Intent-to-treat transmission rates were 3.8% (95% confidence interval: 1.2-8.6) in NVP-NVP, 1.6% (0.2-5.6) in infant-only NVP, and 1.4% (0.4-5.1) in LPV/r arms. As-treated rates were 2.2% (0.5-6.4), 3.2% (0.9-7.9), and 1.5% (0.2-5.2), respectively. Factors independently associated with transmission were prophylaxis duration less than 8 weeks (adjusted odds ratio 15.5; 3.6-66.1) and viral load at baseline at least 4log(10)copies/ml (adjusted odds ratio 10.9; 1.3-91.5). Regimens appeared well tolerated.Conclusion:Transmission rates in all arms were low but noninferiority was not proven. Antiretroviral prophylaxis for at least 8 weeks before delivery is necessary to minimize transmission risk.