Increased gut microbial translocation in HIV-infected children persists in virologic responders and virologic failures after antiretroviral therapy

Abstract

Background: Gut microbial translocation (MT) is considered a major cause of immune activation (IA) and failure of immune reconstitution in HIV infection. This study investigated the relationship of virus replication, IA, CD4 counts and MT in HIV-infected children.

Methods: Lipopolysaccharide, bacterial 16S ribosomal DNA (16SrDNA) and soluble CD14 (sCD14) levels were determined in prospectively collected, stored plasma samples from the Pediatric AIDS Clinical Trials Group Protocol P338, a 48-week study initiated in 1997 to compare efficacy of dual nucleosides with a ritonavir-containing regimen. Results of MT were correlated with study data for T cell IA, plasma viral load and CD4 counts in 85 HIV-infected children (ages 2-17 years) designated as virologic responders or virologic failures (VF) at week 44 based on a cutoff of 400 HIV RNA copies/mL.

Results: Levels of plasma lipopolysaccharide, 16SrDNA and sCD14 were increased in comparison with HIV-uninfected controls and did not decrease at week 44 even in virologic responders. T cell IA was correlated with viral load and sCD14 at entry and with 16SrDNA and sCD14 at week 44 in total patients and in the VF group. Changes in 16SrDNA correlated with changes in IA and negatively with changes in CD4 counts. 16SrDNA was correlated with sCD14 but not with lipopolysaccharide.

Conclusions: MT persists after 44 weeks of antiretroviral therapy in VS and VF patients. In VF, 16SrDNA exhibited relationships to monocyte and T cell IA and CD4 counts but not with viral load, suggesting a dominant role for MT in disease pathogenesis in HIV-infected children.

Figure 1. Plasma levels of microbial products and sCD14 (indicating monocyte activation) are elevated in HIV infected children; LPS does not correlate with 16SrDNA

(A) Correlation between LPS and 16SrDNA; (B, C, D), plasma levels of LPS, 16SrDNA, and sCD14 respectively in total patients; (E), values of sCD14 in virologic responders (VR) and virologic failures (VF). Differences within groups were analyzed using Student’s paired 2-tailed t-test.

Figure 2. Correlation of microbial translocation products with sCD14

(A), Corrleation of LPS with sCD14 at baseline in total patients; (B, C); Correlation of 16SrDNA with sCD14 at week 0 and week 44 respectively in total patients; (D, E), Correlation of 16SrDNA with sCD14 at week 44 in VF and VR groups respectively. Spearman correlations (r and p values) were used to examine the relationships between the markers.

Figure 3. 16sRDNA correlates with T cell immune activation at week 44 in total patients and VF group, but not with VR group

16SrDNA copy number correlated with activated CD4 T cells; CD8 T cells respectively (A, B), in total patients; (C,D), Virologic failures; (E,F), Virologic responders. Spearman correlations (r and p values) were used to examine the relationships between 16SrDNA copy number and activation markers.

Figure 4. Monocyte activation (sCD14) correlates with T cell activation at Week 0 and Week 44 in total patients and in VF, but not in VR patients

Correlation of plasma sCD14 (A,B) with activated CD4 T cells at week 0 and 44 respectively in all patients; (C,D), with activated CD8 T cells at week 0 and 44 in all patients; (E,F), with activated CD4 T cells and activated CD8 T cells respectively at week 44 in VF; (G,H), with activated CD4 T cells and activated CD8 T cells respectively in VR at week 44. Spearman correlations (r and p values) were used to examine the relationships between sCD14 and immune activation markers (percentages of HLA-DR +CD38+) on CD4 and CD8 T cells.

Figure 5. T cell activation correlates with virus load at week 0 in total patients, not at week 44 in total patients or in VF

Virus load correlations with activated CD4 T cells and activated CD8 T cells respectively, (A, B), in total patients at week 0; (C, D), in total patients at week 44; and (E, F) in VF patients at week 44. Spearman correlations (r and p values) were used to examine the relationships between Log HIV RNA and immune activation markers (percentages of HLA-DR +CD38+) on CD4 and CD8 T cells.