Evaluation of vaccine candidates for HIV prevention in Tanzania

Abstract

Background: This thesis describes the capability of an HIV-1 DNA and HIV-1 modified vaccinia virus Ankara (MVA) prime boost vaccination strategy to elicit immune responses known to correlate with reduced risk of HIV infection. Additionally, it describes the confounding effect of vaccine-induced HIV antibodies on the performance of HIV testing algorithms in sub-Saharan Africa, as well as the recruitment and preparation of a cohort of female sex workers (FSW) for potential participation in an HIV vaccine efficacy trial.

Methods: Study I evaluated the capability of the HIV-DNA/MVA vaccine regimen to elicit potent and durable binding antibody responses to the V1V2 domain of HIV-1 gp120. Plasma samples collected at peak immunogenicity and three years later, were analyzed for frequency, magnitude and persistence of antibodies to gp70V1V2 proteins. Study II investigated whether addition of an envelope protein (CN54rgp140/GLA-AF) boost would increase anti-V1V2 responses elicited by the HIV-DNA/HIV-MVA vaccine. Study III assessed the impact of vaccine-induced seroreactivity on the performance of rapid test-based HIV diagnostic algorithms. The HIV diagnostic strategies of Mozambique and Tanzania were evaluated using samples collected from vaccine recipients in the HIVIS and TaMoVac clinical trials. Study IV assessed the suitability of FSW in Dar es Salaam for participation in a phase IIb HIV vaccine efficacy trial (PrEPVacc). HIV incidence, retention rate and risk behaviours were determined in a cohort of 700 FSW after one year of follow up.

Results: Frequent and durable anti-V1V2 responses were elicited in the majority of the vaccine recipients. At peak immunogenicity, 97% of the vaccinees had binding IgG antibodies to the V1V2 loop of CRF01_AE A244. The anti-A244 V1V2 IgG persisted for at least three years in 75% of vaccinees and the response rate was improved by a late HIVMVA boost given three years after the second boost (study I). The CN54rgp140/GLA-AF boost did not enhance the V1V2 response (study II). The HIV diagnostic algorithms in sub- Saharan Africa misdiagnosed a substantial proportion of HIV vaccine recipients. More than half (54%) of the vaccinees would have been incorrectly identified as HIV infected in Tanzania, while, 26.3% would have been misclassified in Mozambique (study III). A high HIV incidence of 3.45 per 100 person years at risk was observed among FSW in the PrEPVacc preparedness study. The rate of HIV acquisition was higher among the young (18-24 years), drug using FSW, and those with syphilis or hepatitis infections (study IV).

Conclusion: Priming with HIV-DNA followed by boosting with HIV-MVA elicited robust and durable anti-V1V2 responses in a majority of vaccinees. Boosting with a combination of CN54rgp140/GLA-AF and HIV-MVA did not augment anti-V1V2 responses elicited by HIV-DNA priming and HIV-MVA boosting. The current HIV testing algorithms in sub- Saharan Africa cannot sufficiently discriminate vaccine-induced seroreactivity from true HIV infection. Young FSW in Dar es Saalam are a suitable target population for HIV vaccine efficacy trials.