Integrase Inhibitor Resistance Mechanisms and Structural Characteristics in Antiretroviral Therapy-Experienced, Integrase Inhibitor-Naive Adults with HIV-1 Infection Treated with Dolutegravir plus Two Nucleoside Reverse Transcriptase Inhibitors in the DAWNING Study

Abstract

At week 48 in the phase IIIb DAWNING study, the integrase strand transfer inhibitor (INSTI) dolutegravir plus 2 nucleoside reverse transcriptase inhibitors demonstrated superiority to ritonavir-boosted lopinavir in achieving virologic suppression in adults with HIV-1 who failed first-line therapy. Here, we report emergent HIV-1 drug resistance and mechanistic underpinnings among dolutegravir-treated adults in DAWNING. Population viral genotyping, phenotyping, and clonal analyses were performed on participants meeting confirmed virologic withdrawal (CVW) criteria on dolutegravir-containing regimens. Dolutegravir binding to and structural changes in HIV-1 integrase-DNA complexes with INSTI resistance-associated substitutions were evaluated. Of participants who received dolutegravir through week 48 plus an additional 110 weeks for this assessment, 6 met CVW criteria with treatment-emergent INSTI resistance-associated substitutions and 1 had R263R/K at baseline but not at CVW. All 7 achieved HIV-1 RNA levels of <400 copies/mL (5 achieved <50 copies/mL) before CVW. Treatment-emergent G118R was detected in 5 participants, occurring with ≥2 other integrase substitutions, including R263R/K, in 3 participants and without other integrase substitutions in 2 participants. G118R or R263K increased the rate of dolutegravir dissociation from integrase-DNA complexes versus wild-type but retained prolonged binding. Overall, among treatment-experienced adults who received dolutegravir in DAWNING, 6 of 314 participants developed treatment-emergent INSTI resistance-associated substitutions, with a change in in vitro dolutegravir resistance of >10-fold and reduced viral replication capacity versus baseline levels. This study demonstrates that the pathway to dolutegravir resistance is a challenging balance between HIV-1 phenotypic change and associated loss of viral fitness. (This study has been registered at ClinicalTrials.gov under identifier NCT02227238.).

Keywords: HIV-1 infection; antiretroviral agents; barrier to resistance; dolutegravir; integrase strand transfer inhibitor.

FIG 1

HIV-1 RNA over time in participants with CVW and INSTI resistance-associated substitutions. Orange circles denote viral load at the week each participant met CVW criteria and also indicate the on-study time point for samples used for population resistance testing and clonal analyses for participants 1 through 3. HIV-1 RNA levels of 400 and 50 copies/mL are indicated by solid and dashed gray lines, respectively. Study visits with HIV-1 RNA <40 copies/mL were plotted as 39 copies/mL. a, the repeat HIV-1 RNA levels were <40 copies/mL at week 24 and 19,486 copies/mL at week 48. b, the repeat HIV-1 RNA level was 3,526 copies/mL at week 36. c, the repeat HIV-1 RNA level was 1,589 copies/mL at week 52. d, the repeat HIV-1 RNA levels were 101 copies/mL at week 52 and 3,011 and 4,146 copies/mL at week 60. e, the repeat HIV-1 RNA levels were 322 copies/mL at week 60 and 6,692 copies/mL at week 72. f, the repeat HIV-1 RNA levels were 203 copies/mL at week 144, 274 copies/mL at week 156, 1,485 copies/mL at week 168, and 12,028 copies/mL at week 180.

FIG 2

Phylogenetic analysis of variant clonal integrase sequences at baseline and CVW. Samples used for clonal analyses at CVW time points are presented in Fig. 1 and are the same samples used for the corresponding population-level resistance data reported in Table 1. Bootstrap confidence levels are indicated on each diagram. CVW, confirmed virologic withdrawal. a, K160T was observed in 4 of 13 clones containing G118R, E138K, and R263K integrase substitutions but is not a prespecified dolutegravir resistance-associated substitution (34).

FIG 3

Variant clone drug sensitivity and integrase region-based replication capacity at baseline and CVW. Replication capacity values for each variant clone are represented by symbol size. Variant clones missing data for replication capacity are indicated by triangles. CVW, confirmed virologic withdrawal; IC₅₀, half-maximal inhibitory concentration. a, values are relative to the wild type. b, the clinical cutoff for dolutegravir is 4.0. c, the biological cutoff for elvitegravir is 2.5. d, the biological cutoff for raltegravir is 1.5. e, values are relative to wild-type replication.

FIG 4

Models of HIV-1 integrase containing wild-type and/or mutant amino acids in the intasome complex. (A) Surface view of the integrase G118R mutant catalytic site (colored by atom with carbons in orange) bound with dolutegravir (rendered in ball-and-stick format and colored by atom with carbons in white). (B) Ribbon-and-stick rendering of the integrase G118R mutant catalytic site (in green) bound with vDNA (in orange). Dolutegravir (as rendered in panel A) binds between the vDNA and tDNA catalytic loops and interacts with the catalytic Mg²⁺ (rendered in ball-and-stick format and colored in green). The tDNA and vDNA catalytic loops are highlighted in pink and magenta, respectively, and illustrate the locations of G118 and N144 (rendered in stick format) on each loop. Hydrogen bonding interactions between N144 with R263 and G118R with E92 are indicated by dashed yellow lines. (C) The integrase G118R mutant catalytic site (as rendered and oriented in panel B) bound to both vDNA and tDNA substrates (in orange). Hydrogen bonding interactions among G118R, the 3′ terminus of the tDNA, and the Mg²⁺ are indicated by dashed yellow lines. (D) Ribbon-and-stick rendering of the integrase catalytic site containing R263K (colored in magenta) and wild-type G118 bound with vDNA and tDNA substrates. DTG, dolutegravir; tDNA, host target DNA; vDNA, viral DNA.