10-1074 and 3BNC117 are broadly neutralising antibodies (bNAbs), which target the V3 glycan and the CD4bs respectively, and together can maintain viral suppression after antiretroviral treatment (ART) interruption. Due to its high diversity rate and the dense array of glycans that shield the underlying bNAb epitopes, HIV can escape neutralisation. Although there is no established bNAb sensitivity screening method, several algorithms have identified certain genetic signatures that may predict potential bNAb sensitivity. Here, we aim to assess the utility of bNAb sensitivity screening for clinical trials and to present the distribution of 10-1074 and 3BNC117 sensitivity landscape in a UK cohort.
Samples from 173 participants diagnosed and treated during primary HIV infection (PHI), within an estimated 6 months of seroconversion were processed. All participants had been on ART for >1 year and had undetectable viral load at the time of sampling. An average of 20 proviral env sequences per sample was amplified using single genome amplification from 148 participants. Following sequencing, we inspected the amino acid residues that have been reported to confer resistance to 10-1074 (N332 glycosylation motif and [sup]324[/sup]GDIR[sup]327[/sup]) and 3BNC117 (positions D279, N280 and [sup]456[/sup]RDGG[sup]459[/sup]) epitopes.
A total of 3138 proviral env sequences, mainly B clade (70.9%), were sequenced and analysed (Table). Mutations associated with resistance to either or both bNAbs were detected in 47.9% of participants and notably, 36.6% of these contained a mixture of both resistant and sensitive sequences. 66.1% of participants with resistant sequences had 10-1074 associated mutations. Mutations affecting the N332 glycosylation motif Asn-X-Ser/Thr were the most common 10-1074 resistance-associated mutations (85%). The most frequently mutated 3BNC117 sites were 456 and 459 (47.8% and 34.7%, respectively). There was however considerable variation between participants, including between those with mixed and full resistance. Phylogenetic analysis suggested evidence for both transmitted resistance and in-host evolution.
Our findings show that around half the cohort treated during PHI has potential pre-existing resistance to 10-1074 and 3BNC117 based on current algorithms. Although it is unclear how well these algorithms predict clinical response to bNAbs in real world settings, the suggestion from these data is that screening may be key to guide effective treatment.