Background
Antiretroviral therapy fails to eliminate a reservoir of HIV-1 infected cells wherein the provirus is transcribed at levels insufficient to induce cell death. Earlier studies of broadly neutralizing antibodies (bNAbs) administered during ART interruption (ATI) showed mixed results with some reporting accelerated reservoir decay. RIO enabled examination of reservoir dynamics in 32 individuals receiving 3BNC117-LS and 10-1074-LS during ATI or after ART restart.
Methods
Participants were randomized 1:1 to bNAbs or placebo and underwent ATI. After initial rebound, the placebo group received bNAbs at ART restart. Intact and defective reservoirs were measured by digital droplet PCR (ddPCR) and by full genome sequencing (Q4PCR) at baseline (before ATI) and after an average of 56 weeks after the initial bNAb dose at a time when HIV-1 was undetectable.
Results
Individuals in the 2 arms showed similar numbers of intact and defective proviruses. Consistent with early therapy, 5 out of 27 and 7 out of 21 individuals in bNAb and placebo arms respectively, had no measurable reservoir at baseline by ddPCR. The geometric mean number of intact proviruses/106 CD4+T cells was 8.4 by ddPCR and 0.28 by Q4PCR. There was no significant correlation between reservoir size and time to initial rebound in the placebo arm. 80% of the CD4+T cells carrying intact proviruses were members of expanded clones. 267 representative proviral Env genes from reservoir and rebound were produced as pseudoviruses, and subjected to sensitivity testing in TZM-bl assays. All individuals in the bNAb arm, and all but 1 in the placebo arm showed reservoir sensitivity to at least 1 of the 2 bNAbs. Despite low to non-measurable autologous antibody neutralizing activity to reservoir viruses in the placebo group, rebound sequences only corresponded to reservoir sequences in 2 of 9 individuals. Therefore, autologous antibody neutralizing activity cannot account for the disparity between intact proviruses found in the reservoir and rebound in these individuals. Reservoir half-life and log fold change after bNAb dose during ATI (bNAb arm) or at ART restart (placebo arm) were not significantly different. When all measurements for both arms are considered jointly (n=32) the half-life of intact and defective reservoirs were 33 and 424 weeks respectively.
Conclusions
Antibody therapy in the RIO trial was associated with an average intact reservoir half-life decay which is 7-8 fold faster than prior reports.