SummaryBackgroundModest efficacy was reported for the HIV vaccine tested in the RV144 trial, which comprised a canarypox vector (ALVAC) and envelope (env) glycoprotein (gp120). These vaccine components were adapted to express HIV-1 antigens from strains circulating in South Africa, and the adjuvant was changed to increase immunogenicity. Furthermore, 12-month immunisation was added to improve durability. In the HIV Vaccine Trials Network (HVTN) 100 trial, we aimed to assess this new regionally adapted regimen for advancement to efficacy testing.MethodsHVTN 100 is a phase 1/2, randomised controlled, double-blind trial at six community research sites in South Africa. We randomly allocated adults (aged 18–40 years) without HIV infection and at low risk of HIV infection to either the vaccine regimen (intramuscular injection of ALVAC-HIV vector [vCP2438] at 0, 1, 3, 6, and 12 months plus bivalent subtype C gp120 and MF59 adjuvant at 3, 6, and 12 months) or placebo, in a 5:1 ratio. Randomisation was done by computer-generated list. Participants, investigators, and those assessing outcomes were masked to random assignments. Primary outcomes included safety and immune responses associated with correlates of HIV risk in RV144, 2 weeks after vaccination at 6 months (month 6·5). We compared per-protocol participants (ie, those who completed the first four vaccinations and provided samples at month 6·5) from HVTN 100 with stored RV144 samples assayed contemporaneously. This trial is registered with the South African National Clinical Trials Registry (DOH-27-0215-4796) and ClinicalTrials.gov (NCT02404311).FindingsBetween Feb 9, 2015, and May 26, 2015, 252 participants were enrolled, of whom 210 were assigned vaccine and 42 placebo. 222 participants were included in the per-protocol analysis (185 vaccine and 37 placebo). 185 (100%) vaccine recipients developed IgG binding antibodies to all three vaccine-matched gp120 antigens with significantly higher titres (3·6–8·8 fold; all p<0·0001) than the corresponding vaccine-matched responses of RV144. The CD4+ T-cell response to the ZM96.C env protein in HVTN 100 was 56·4% (n=102 responders), compared with a response of 41·4% (n=79 responders) to 92TH023.AE in RV144 (p=0·0050). The IgG response to the 1086.C variable loops 1 and 2 (V1V2) env antigen in HVTN 100 was 70·5% (95% CI 63·5–76·6; n=129 responders), lower than the response to V1V2 in RV144 (99·0%, 95% CI 96·4–99·7; n=199 responders).InterpretationAlthough the IgG response to the HVTN 100 vaccine was lower than that reported in RV144, it exceeded the predicted 63% threshold needed for 50% vaccine efficacy using a V1V2 correlate of protection model. Thus, the subtype C HIV vaccine regimen qualified for phase 2b/3 efficacy testing, a critical next step of vaccine development.FundingUS National Institute of Allergy and Infectious Diseases (NIAID), and Bill & Melinda Gates Foundation.

Background The Phambili study, conducted in South Africa amongst a predominantly heterosexual population, evaluated the efficacy of the MRK Ad5 gag/pol/nef subtype B HIV-1 preventive vaccine. Enrollment and vaccinations were stopped, participants unblinded, and follow-up extended when the Step study evaluating the same vaccine in the Americas, Caribbean and Australia was unblinded for non-efficacy with more HIV infections amongst vaccinee than placebo recipients [ZM1]. Extensive analyses over the complete follow-up period, most of which was unblinded, are reported. Methods Phambili participants were HIV-1 uninfected, sexually active men and women aged 18–35 years, followed for 3.5 years. HIV testing and risk reduction counseling occurred at weeks 0, 12, 30 and were switched to a 3 monthly schedule after unblinding. Cox proportional hazards models were used to estimate HIV-1 infection hazard ratios (HR) comparing vaccine to placebo recipients, overall and within subgroups. Long-term vaccine efficacy was evaluated in participants who were unblinded early in follow-up. Results Of the 801 participants enrolled (400 vaccine, 401 placebo), 112 (28%) received 1 vaccination, 259 (65%) 2 vaccinations and 29(7%) 3 vaccinations. More infections occurred in vaccinees (n=63) as compared to placebo (n=37) (adjusted HR (vaccine:placebo) 1.70, 95% CI 1.13–2.55, p = 0.01). We found no increase in infections with the number of vaccinations received and that the HRs did not differ by gender, circumcision, or Ad5 serostatus. Differences in risk behavior at baseline or during the study, or differential drop-out (p=0.40) are unlikely explanations for the increased rate of HIV-1 infections seen in vaccinees. Conclusion The increased HR of HIV-1 acquisition, irrespective of number of doses received, warrants further investigation to understand the biological mechanism. Further use of the Ad5 vector for HIV vaccines is not warranted

No abstract