Abstract Body

Background: Rilpivirine/emtricitabine/tenofovir DF (RPV/FTC/TDF) is currently approved for use in HIV-1 treatment-naïve adults with HIV-1 RNA ≤100,000 c/mL due to increased virologic failure with resistance development in patients with high baseline viral load (BL VL). At Week 96 in the phase 3b STaR study (GS-US-264-0110; RPV/FTC/TDF vs. efavirenz/FTC/TDF in treatment-naïve adults with no BL VL restriction), 5.3% of patients on RPV/FTC/TDF developed resistance mutations by population sequencing (9.0% BL VL >100,000 c/mL vs. 3.5% BL VL ≤100,000 c/mL). Deep sequencing was performed to assess the potential clinical impact of pre-existing minority variants in RPV/FTC/TDF-treated patients with high BL VL.

Methods: Deep sequencing (Illumina MiSeq) was performed on baseline and virologic failure (VF) samples for 24 RPV/FTC/TDF-treated patients in the resistance analysis population (RAP; patients with HIV-1 RNA ≥400 c/mL at VF, discontinuation, or Weeks 48 or 96). Baseline samples from 44 non-RAP RPV/FTC/TDF-treated BL VL and BL CD4 matched control patients, including all 29 with BL VL ≥500,000 c/mL, were also analyzed. Deep sequencing results (≥1% cutoff) were compared to population sequencing.

Results: Baseline NRTI- or NNRTI-associated mutations were detected in 29% (7/24) of RAP isolates and 27% (12/44) of controls by deep sequencing. Study exclusion mutations were detected at low frequencies in 4 RAP patients (BL VL >100,000 c/mL: E138K, H221Y n=1 each; BL VL ≤100,000 c/mL: M184V, E138G n=1 each). While all 4 patients developed NRTI- and/or NNRTI-associated mutations in their HIV-1 at VF, none of these same mutations were detected except for 1 patient with 1.17% M184V at baseline who had >99% M184V at VF. Exclusion mutations were also detected at low frequencies in 7 control patients (BL VL >100,000 c/mL: K65R, E138K n=2 each; Y181C, E138G n=1 each; BL VL ≤100,000 c/mL: E138G n=1). Six were virologic successes at Week 96; 1 patient with 1.08% K65R discontinued at Week 32 with HIV-1 RNA 64 c/mL. Additional NRTI- or NNRTI-associated mutations not detected by population sequencing at VF were detected in 41% (9/22) of RAP VF isolates with deep sequencing data.

Conclusions: The presence of low-frequency NRTI- and NNRTI-associated mutations at baseline was not predictive for VF or resistance development for RPV/FTC/TDF-treated patients in the STaR study. Deep sequencing results were generally consistent with population sequencing but detected additional low-frequency resistance mutations at VF.