Background:
Waning broadly neutralizing antibody (bNAb) levels and emergence of resistance have been associated with viral rebound during analytic treatment interruption (ATI) studies. In this pharmacokinetic/pharmacodynamic (PK/PD) analysis, we evaluated the impact of bNAb exposure, susceptibility, and antidrug antibody (ADA) formation on rebound kinetics following combination immunotherapy with a boosted DNA vaccine, lefitolimod, and bNAbs (10-1074 and VRC07-523LS) (NCT04357821), in which 7/10 individuals exhibited altered post-intervention rebound dynamics.
Methods:
We described plasma bNAb PK using population PK modeling approaches in Monolix software. We performed Spearman correlations or Wilcox’s test to determine the relationship between bnAb AUC, peak concentration (Cmax), and bNAb level at time of rebound and rebound kinetics (time to rebound, post ATI setpoint). We evaluated susceptibility (IC90) relative to bNAb level (i.e., IQ90: bNAb level/IC90). We performed competitive and functional ADA assays for both bNAbs.
Results:
Time to Rebound: Greater bNAb exposure for 10 participants (9 cis men, 1 trans woman) was associated with later rebound (VRC07-523LS AUC ρ=0.76, p=0.04; 10-1074 Cmax ρ =0.7, p=0.04). In those who rebounded later (>15 weeks), VRC07-523LS and 10-1074 levels were lower at the time of rebound (both p=0.01), with a similar trend observed for IQ90 (p=0.01 and p=0.03, for each bNAb). There was no association between IC90 and time to rebound. Post intervention setpoints: Although there was no association between VRC07-523LS IC90 and setpoint, higher VRC07-523LS IQ90 at the time of rebound was associated with higher post-intervention setpoint (ρ = 0.85, p=0.006). No association between 10-1074 IQ90 or IC90 and setpoint were observed. Antidrug antibody: While ADA was detected for two participants for competitive assays for both bNAbs, no functional ADA impacting PK was observed.
Conclusions:
Higher bNAb exposure (e.g., AUC, Cmax) was consistently associated with delayed viral rebound. Our results suggest that post-treatment setpoint was not driven by bNAb susceptibility and that the association of IQ90 and setpoint is driven by higher VRC07-523LS levels at the time of rebound in those who rebounded earlier and had higher setpoints. Overall, bNAb PK-PD is likely not responsible for lower observed post-treatment setpoints during this trial, suggesting the effect is likely attributable to changes in anti-HIV immune function.
