HIV-1 phenotypic reverse transcriptase inhibitor drug resistance test interpretation is not dependent on the subtype of the virus backbone

Abstract

To date, the majority of HIV-1 phenotypic resistance testing has been performed with subtype B virus backbones (e.g. HXB2). However, the relevance of using this backbone to determine resistance in non-subtype B HIV-1 viruses still needs to be assessed. From 114 HIV-1 subtype C clinical samples (36 ARV-naïve, 78 ARV-exposed), pol amplicons were produced and analyzed for phenotypic resistance using both a subtype B- and C-backbone in which the pol fragment was deleted. Phenotypic resistance was assessed in resulting recombinant virus stocks (RVS) for a series of antiretroviral drugs (ARV's) and expressed as fold change (FC), yielding 1660 FC comparisons. These Antivirogram® derived FC values were categorized as having resistant or sensitive susceptibility based on biological cut-off values (BCOs). The concordance between resistance calls obtained for the same clinical sample but derived from two different backbones (i.e. B and C) accounted for 86.1% (1429/1660) of the FC comparisons. However, when taking the assay variability into account, 95.8% (1590/1660) of the phenotypic data could be considered as being concordant with respect to their resistance call. No difference in the capacity to detect resistance associated with M184V, K103N and V106M mutations was noted between the two backbones. The following was concluded: (i) A high level of concordance was shown between the two backbone phenotypic resistance profiles; (ii) Assay variability is largely responsible for discordant results (i.e. for FC values close to BCO); (iii) Confidence intervals should be given around the BCO's, when assessing resistance in HIV-1 subtype C; (iv) No systematic resistance under- or overcalling of subtype C amplicons in the B-backbone was observed; (v) Virus backbone subtype sequence variability outside the pol region does not contribute to phenotypic FC values. In conclusion the HXB2 virus backbone remains an acceptable vector for phenotyping HIV-1 subtype C pol amplicons.

Figure 1. Process flow diagram for phenotyping procedure within subtype B- and C- backbones.

Figure 2. Plots showing discordant phenotypic resistance comparisons, and variation around the BCO.

The FC of the B backbone derived viruses (assumed to be the gold standard) are shown on the x-axis, and the difference in log FC values (B backbone FC – C backbone FC) on the y-axis. The shaded region on the plots is the region wherein the log FC biological cut-offs lie for the 18 ARVs tested. The mean +2SD and mean −2SD are drawn in to illustrate the natural variation around the BCO. Any point s found above the mean +2SD or below the mean −2SD are considered to be truly discordant. Figure 2A is a plot of comparisons within the B'sens/C'res group. Figure 2B is a plot of comparisons within the B'res/C'sens group.