Background:
BnAbs are being tested in clinical trials of treatment and cure in people with HIV (PWH) who have existing autologous neutralizing antibody (anAb) responses. To determine the relative selective pressure of bnAbs and anAbs in the BEAT2 study of 3BNC117, 10-1074, and IFNa2b, we performed a sieve analysis comparing the potency of administered bnAbs and host anAbs against reservoir and rebound Envs.
Methods:
In 8 participants of BEAT2 (NCT03588715), we sequenced provirus from PBMCs collected at study enrollment via FLIPseq or MIPseq (n= 314) to identify intact HIV-1 genomes in single and clonally-expanded populations. In 12 participants, plasma rebound envs (n=233) were sequenced by SGS at first detectable rebound viremia. Reservoir and rebound Envs were cloned and tested for neutralization sensitivity via TZM.bl assay to bnAbs and longitudinal plasma IgG (collected at a median of 8 time points over 12-24 study months). Statistical comparisons were by Wilcoxon matched-pairs test.
Results:
In all 8 participants, rebound envs aligned within reservoir phylogenies, but were not identical to any sampled reservoir viruses. Comparing rebound Envs (1 per participant) and reservoir Envs (2-6 per participant), rebound Envs were significantly more resistant to 10-1074 (IC50 6.44 µg/ml vs. 0.71 µg/ml; p=0.02), and trended towards greater resistance to 3BNC117 (ns). When tested against baseline plasma IgG (prior to bnAb dosing), rebound Envs were generally more resistant than reservoir Envs, with markedly greater resistance (30-fold difference in IC50) in the 2 participants with most delayed rebound (>12 weeks post-bnAb dosing). Potency of plasma IgG against both rebound and reservoir Envs rose significantly during bnAb dosing (mean >3-fold change in IC50 for both; p=0.001). Post-ART restart, after bnAbs had waned and anAbs had responded to recent rebound viremia, plasma IgG potency increased vs. rebound Envs (mean 2.5-fold change in IC50; p=0.001), but not reservoir Envs (ns).
Conclusions:
In the BEAT2 study of 2 bnAbs and IFNa2b, the greater potency of the administered bnAbs against reservoir vs. rebound Envs indicates bnAb selective pressure. In 2 participants with delayed rebound, baseline anAbs also exerted selective pressure. After rebound and ART restart, anAbs evolved to selectively target rebound Envs. Together, results suggest that anAbs contributed to virus suppression in a subset (25%) of studied bnAb trial participants. Approaches to boost anAbs may increase this proportion.