Phenotypic characterization of virological failure following lopinavir/ritonavir monotherapy using full-length Gag-protease genes

Abstract

Objectives: Major protease mutations are rarely observed following first-line failure with PIs and interpretation of genotyping results in this context may be difficult. We performed extensive phenotyping of viruses from five patients failing lopinavir/ritonavir monotherapy in the MONARK study without major PI mutations by standard genotyping.

Methods: Phenotypic susceptibility testing and viral infectivity assessments were performed using a single-cycle assay and fold changes (FC) relative to a lopinavir-susceptible reference strain were calculated.

Results: >10-fold reduced baseline susceptibility to lopinavir occurred in two of five patients and >5-fold in another two. Four of five patients exhibited phylogenetic evidence of a limited viral evolution between baseline and failure, with amino acid changes at drug resistance-associated positions in one: T81A emerged in Gag with M36I in the protease gene, correlating with a reduction in lopinavir susceptibility from FC 7 (95% CI 6-8.35) to FC 13 (95% CI 8.11-17.8). Reductions in darunavir susceptibility (>5 FC) occurred in three individuals.

Discussion: This study suggests both baseline reduced susceptibility and evolution of resistance could be contributing factors to PI failure, despite the absence of classical PI resistance mutations by standard testing methods. Use of phenotyping also reveals lower darunavir susceptibility, warranting further study as this agent is commonly used following lopinavir failure.

Keywords: Gag; HIV; antiretroviral resistance; monotherapy; protease inhibitors.

Figure 1.

Patient sample selection flow diagram.

Figure 2.

Phenotypic PI susceptibility. Phenotypic susceptibility to four PIs atazanavir (ATV), darunavir (DRV), lopinavir (LPV) and saquinavir (SQV) was determined for viral variants derived from five patients: (a) 1403, (b) 3204, (c) 1404, (d) 4201 and (e) 508. Data are presented as FC in EC₅₀ in comparison with the assay reference strain. BL, baseline; F, failure.

Figure 3.

Maximum likelihood phylogeny. A maximum likelihood tree constructed using the GTR model in PhyML using an alignment of all variants from baseline and failure timepoints from five patients. Variants from each patient are represented by a single colour, the screening variants by triangles and those from failure with circles. Nodes supported by >75% bootstrapping (≥350/500) are marked by an asterisk (*). This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.