Effects of Combined CCR5/Integrase Inhibitors-Based Regimen on Mucosal Immunity in HIV-Infected Patients Naïve to Antiretroviral Therapy: A Pilot Randomized Trial

Abstract

Whether initiation of antiretroviral therapy (ART) regimens aimed at achieving greater concentrations within gut associated lymphoid tissue (GALT) impacts the level of mucosal immune reconstitution, inflammatory markers and the viral reservoir remains unknown. We included 12 HIV- controls and 32 ART-naïve HIV patients who were randomized to efavirenz, maraviroc or maraviroc+raltegravir, each with fixed-dose tenofovir disoproxil fumarate/emtricitabine. Rectal and duodenal biopsies were obtained at baseline and at 9 months of ART. We performed a comprehensive assay of T-cell subsets by flow cytometry, T-cell density in intestinal biopsies, plasma and tissue concentrations of antiretroviral drugs by high-performance liquid chromatography/mass spectroscopy, and plasma interleukin-6 (IL-6), lipoteichoic acid (LTA), soluble CD14 (sCD14) and zonulin-1 each measured by ELISA. Total cell-associated HIV DNA was measured in PBMC and rectal and duodenal mononuclear cells. Twenty-six HIV-infected patients completed the follow-up. In the duodenum, the quadruple regimen resulted in greater CD8+ T-cell density decline, greater normalization of mucosal CCR5+CD4+ T-cells and increase of the naïve/memory CD8+ T-cell ratio, and a greater decline of sCD14 levels and duodenal HIV DNA levels (P = 0.004 and P = 0.067, respectively), with no changes in HIV RNA in plasma or tissue. Maraviroc showed the highest drug distribution to the gut tissue, and duodenal concentrations correlated well with other T-cell markers in duodenum, i.e., the CD4/CD8 ratio, %CD4+ and %CD8+ HLA-DR+CD38+ T-cells. Maraviroc use elicited greater activation of the mucosal naïve CD8+ T-cell subset, ameliorated the distribution of the CD8+ T-cell maturational subsets and induced higher improvement of zonulin-1 levels. These data suggest that combined CCR5 and integrase inhibitor based combination therapy in ART treatment naïve patients might more effectively reconstitute duodenal immunity, decrease inflammatory markers and impact on HIV persistence by cell-dependent mechanisms, and show unique effects of MVC in duodenal immunity driven by higher drug tissue penetration and possibly by class-dependent effects.

Fig 1. Study profile.

Fig 2. Changes in peripheral and duodenal CD4⁺ and CD8⁺ T-cells.

Panel A. Changes in peripheral CD4⁺ T-cell counts. Panel B. Changes in peripheral CD4/CD8 ratio. Panel C. Changes in duodenal CD4⁺ T-cell density. Panel D. Changes in duodenal CD8⁺ T-cell density. The purple dash line represents cross-sectional mean values in controls. The equality of all mean changes among the three groups was tested using a Wald test for the interaction term and is referred as “Overall differences”. A statistically significant interaction term (group*month) identified significantly different slopes between treatment groups.

Fig 3. Changes in the naive/memory ratio and CD38⁺HLA-DR⁺ naive CD8 T-cells in duodenum.

Panel A. Changes in the naïve/memory ratio in duodenum for CD4⁺ (I).and CD8⁺ T-cells (II). Panel B. Changes in %HLA-DR⁺CD38⁺ /naive CD8 T-cells in blood (I), rectum (II) and duodenum (III). The purple dash line represents cross-sectional mean values in controls. The equality of all mean changes among the three groups was tested using a Wald test for the interaction term and is referred as “Overall differences”. A statistically significant interaction term (group*month) identified significantly different slopes between treatment groups.

Fig 4. Correlations between the %naive-activated CD8⁺ T-cells and the naïve/memory CD8 T-cell ratio in blood, rectum and duodenum.

Fig 5. Changes in %CCR5 T-cells in blood, rectum and duodenum.

Panel A. Changes in %CD4⁺CCR5⁺ T-cells in blood. Panel B. Changes in %CD4⁺CCR5⁺ T-cells in rectum. Panel C. Changes in %CD4⁺CCR5⁺ T-cells in duodenum. The Purple dash line represents cross-sectional mean values in controls. The equality of all mean changes among the three groups was tested using a Wald test for the interaction term and is referred as “Overall differences”. A statistically significant interaction term (group*month) identified significantly different slopes between treatment groups.

Fig 6. Effects of three ART regimens on markers of viral persistence.

Panel A. HIV RNA in plasma. Panel B. Cell-associated HIV RNA in rectum. Panel C. Cell-associated HIV RNA in duodenum. Panel D. Cell-associated HIV DNA in PBMC. Panel E. Cell-associated HIV DNA in rectum. Panel F. Cell-associated HIV DNA in duodenum. The equality of all mean changes among the three groups was tested using a Wald test for the interaction term and is referred as “Overall differences”. A statistically significant interaction term (group*month) identified significantly different slopes between treatment groups.

Fig 7. Tissue drug distribution.

Panels A-B. Percentages of rectum/plasma (A) and duodenum/plasma (B) drug concentrations. Maraviroc reached the highest distribution to rectum and duodenum (all P<0.005). Panel C-E. Correlations between plasma, rectum and duodenal levels of EFV (C), MVC (D) and RAL (E). MVC plasma levels correlated better with tissue levels than RAL or EFV. The Purple dash line represents cross-sectional mean values in controls.

Fig 8. Correlations between MVC duodenal concentrations and several markers of duodenal immunity.

Fig 9. Absolute concentrations in each compartment of efavirenz (A), maraviroc (B) and raltegravir (C) related to the IC50 for HIV-1 replication.

Fig 10. Effects of three ART regimens on markers of innate immunity.

Panel A. Changes in plasma IL-6. Panel B. Changes in plasma sCD14. Panel C. Changes in plasma LTA. Panel D. Changes in plasma zonulin-1. The Purple dash line represents cross-sectional mean values in controls. The equality of all mean changes among the three groups was tested using a Wald test for the interaction term and is referred as “Overall differences”. A statistically significant interaction term (group*month) identified significantly different slopes between treatment groups.