Background:  Among untreated HIV-infected individuals, CXCR4 (X4)-tropic viruses are uncommon except in advanced stages of immunodeficiency. However, little is known about the prevalence and immunologic consequences of X4 tropism in treated patients with detectable viremia.

Methods:  The chemokine receptor tropism of plasma viral isolates was determined by the HIV Entry Assay and compared between untreated and ART-treated, chronically infected patients sampled from 2 San Francisco cohorts. Differences between groups were adjusted for CCR5  d32 genotype, nadir CD4⁺ T cell count, and duration of HIV infection. The association between tropism and naïve (CD45RA⁺CD62L⁺), activated (HLA-DR⁺CD38⁺), and non-activated memory CD4 counts was also assessed.

Results:  Both the 81 untreated patients and the 186 treated patients were similar as to median CD4 counts (258 vs 294 cells/mm³) and years since initial HIV diagnosis (13 vs 12 years), but untreated patients had lower median plasma HIV RNA levels (3.6 vs 4.0 log₁₀ copies/mL) and nadir CD4 counts (60 vs 203 cells/mm³). The treated patients had a median of 4 nucleoside reverse transcriptase inhibitor (NRTI)-associated and 2 major protease inhibitor (PI)-associated mutations. Of 186 treated patients, 75 (40%) harbored dual/mixed or X4-tropic viruses compared with only 12 of 81 (15%) untreated participants, p <0.001. Among all participants, dual/mixed or X4 tropism was more common in those with lower current and nadir CD4 counts (p <0.001 for both comparisons) and in those heterozygous for the CCR5 d32 polymorphism (p = 0.02). Even after adjustment for nadir CD4 count, duration of HIV infection, and CCR5 d32 genotype, treated patients had 4-fold greater odds of dual/mixed or X4 tropism than untreated patients, p = 0.004. Patients harboring dual/mixed or X4-tropic viruses had lower naïve (p = 0.05) and resting memory CD4 counts (p = 0.02) than those harboring R5-tropic viruses, but similar activated CD4 counts (p = 0.27). Furthermore, after adjustment for plasma HIV RNA levels, treated patients harboring dual/mixed or X4 tropic viruses were maintaining 78 fewer CD4⁺ T cells/mm³ above their pre-treatment nadir than those harboring R5-tropic viruses (p = 0.002).

Conclusions:  Treated patients with partial viral suppression are more likely than untreated patients to harbor dual/mixed or X4-tropic viruses, independent of the extent of current or prior immunodeficiency. Dual/mixed or X4 tropism is also associated with fewer treatment-mediated CD4⁺ T cell gains, perhaps due to a greater ability to deplete resting memory and naïve CD4 cells.