Background:  HIV tropism has important implications in the pathogenesis of HIV infection. Around 15% of HIV-1 subtype B ART-naïve, chronically infected patients carry X4 or X4/R5 mixed-dual tropic viruses. This prevalence may increase as much as 40% in patients with late disease stage failing antiretroviral therapy. Preliminary reports have suggested that HIV-1 subtypes A and D might be more prone to show X4 viruses. Information regarding tropism for other HIV-1 clades is scarce.

Methods:  Individuals known to be infected with HIV-1 non-B subtypes at our institution were identified. A control group of subjects infected with clade B viruses was examined in parallel. HIV tropism was evaluated using a new phenotypic assay (PHENOSCRIPT®ENV), which is based on the recombination of gp120-gp41 PCR products in a plasmid. V3 loop amino acid sequences derived from plasma RNA from each individual were used to estimate co-receptor usage using a phenotype predictor software (http://genomiac2.ucsd.edu:8080 /wetcat/v3.html). Viral load, CD4 counts and antiretroviral therapy at the time of the analysis were recorded.

Results:  We analyzed 102 HIV-1⁺ individuals (67 non-B and 35 B subtypes). The distribution of HIV subtypes was:  7A, 35B, 7C, 1D, 1F, 16G, 1H, 1J, 3CRF01_AE, 19CRF02_AG, 3CRF12_BF, 2CRF14_BG, and 6URF. The overall prevalence of X4 or X4/R5 mixed-dual tropic viruses was 18.7% (11.8% in drug-naïve and 26.8% in ART-experienced patients). No significant differences were observed comparing B and non-B subtypes after adjusting for antiretroviral exposure. Individuals with X4 viruses had significant lower mean CD4 counts compared to R5 patients (177 vs 343; p = 0.035), but there were no significant differences in mean plasma viremia. The overall concordance between tropism results assessed using the phenotypic and genotypic tests was 80.6%, without significant differences between B and non-B clades. However, a higher discordance was noticed for CRF02_AG specimens (30%). Overall, the genotypic tool tended to overestimate X4 viruses with respect to the phenotypic assay, especially in antiretroviral-experienced patients infected with clade B viruses.

Conclusions:  The overall prevalence of X4 viruses seems to be similar in clade B and non-B viruses. The performance of the new PHENOSCRIPT®ENV assay testing non-B subtypes was good. In contrast, tropism estimations based on V3 genotypic scores might face some limitations confronting some non-B subtypes.