Background:  A complex latent reservoir of previously selected HIV variants persists during combination antiretroviral therapy. Interruption of therapy in patients with drug-resistant HIV is often associated with the emergence of these archived viruses as well as rapid losses in CD4⁺ T-cell counts. The degree to which these dynamic changes reflect changes in HIV co-receptor tropism (R5 vs X4-tropic) has not been defined.

Methods:  We analyzed longitudinally collected data from 28 individuals who had detectable drug-resistant HIV on a stable regimen and who experienced a shift to drug-susceptible HIV during a structured treatment interruption. Samples were selected which corresponded to a change in drug-resistant phenotype. Envelope of HIV-1 was amplified from patient plasma samples and recombinant viruses were used to infect cells expressing either CCR5 or CXCR4. Tropism was determined by measuring luciferase activity in the presence of R5 or X4 inhibitors. Gp 160 envelopes were sequenced as virus populations and as individual envelope clones.

Results:  Longitudinal tropism data are available on 20 subjects. Prior to therapy interruption, the dominant plasma virus was exclusively R5 tropic in 10 subjects and either dual/mixed or X4 tropic in 10 subjects. Concurrent with emergence of drug-susceptible “wild-type” (WT) HIV, the virus population in plasma shifted co-receptor utilization in 5 subjects. There were no consistent trends in the direction of this shift (2 shifted from dual topic to either R5 or X4; 2 from R5 to dual tropic; and one from X4 to dual tropic). Genetic changes during the tropism switch were observed in the V3 loop. We next assessed the impact of baseline tropism (pure R5 vs R5/X4) on change in CD4⁺ T-cell count using repeated measures analysis. A total of 288 CD4 observations in 28 subjects interrupting therapy contributed to this analysis. After controlling for baseline CD4 and change in viral load, we observed no evidence of difference in CD4 declines between those with pure R5 vs those with dual-tropic virus (P = 0.20).

Conclusions:  Patients virologically failing combination antiretroviral therapy may harbor viruses with a tropism that differs from that in plasma. These viruses may impact on response to subsequent R5 inhibitor-based salvage therapy. The relative impact of R5 vs X4 on CD4 T-cell counts will require careful observation in larger cohorts.

Keywords: Chemokine receptor tropism; viral fitness; structured treatment interruption