Transient expansion of activated CD8(+) T cells characterizes tuberculosis-associated immune reconstitution inflammatory syndrome in patients with HIV: a case control study

Abstract

Background: CD4(+) T cell activation indicators have been reported to be a common phenomenon underlying diverse manifestations of immune reconstitution inflammatory syndrome (IRIS). However, we have found that a high frequency of circulating CD8(+) T cells is a specific risk factor for mycobacterial IRIS. Therefore, we investigated whether CD8(+) T cells from patients who develop TB IRIS were specifically activated.

Methods: We obtained PBMCs from HIV+ patients prior to and 4, 8, 12, 24, 52 and 104 weeks after initiating antiretroviral therapy. CD38 and HLADR expression on naive, central memory and effector memory CD8(+) and CD4(+) T cells were determined by flow cytometry. Absolute counts and frequencies of CD8(+) T cell subsets were compared between patients who developed TB IRIS, who developed other IRIS forms and who remained IRIS-free.

Results: TB IRIS patients showed significantly higher counts of naive CD8(+) T cells than the other groups at most time points, with a contraction of the effector memory subpopulation occurring later in the follow-up period. Activated (CD38(+) HLADR(+)) CD8(+) T cells from all groups decreased with treatment but transiently peaked in TB IRIS patients. This increase was due to an increase in activated naive CD8(+) T cell counts during IRIS. Additionally, the CD8(+) T cell subpopulations of TB IRIS patients expressed HLADR without CD38 more frequently and expressed CD38 without HLADR less frequently than cells from other groups.

Conclusions: CD8(+) T cell activation is specifically relevant to TB IRIS. Different IRIS forms may involve different alterations in T cell subsets, suggesting different underlying inflammatory processes.

Keywords: Activation; CD8 T cells; HIV-1; HIV-2; Highly active anti-retroviral therapy (HAART); Human immunodeficiency virus (AIDS); Inflammation.

Figure 1

Absolute counts of CD8 T cell subpopulations during antiretroviral treatment. Absolute counts of naive (A), central memory (B), and effector memory (C) CD8 T cells before (week 0) and during antiretroviral treatment in each patient group. Values correspond to mean count ± 1 SEM. * Significant difference between the TB IRIS and No IRIS groups. # Significant difference between the TB IRIS and Other IRIS groups. Two-group differences were determined only when the Kruskal-Wallis test showed overall group effects.

Figure 2

Different patterns of CD38 and HLADR expression within CD8⁺ T cell maturation subpopulations. Percentage of CD38⁺ HLADR^- cells among naive (A), CM (B), and EM (C) CD8⁺ T cells. Percentage of CD38⁺ HLADR⁺ cells among naive (D), CM (E), and EM (F) CD8 T cells throughout the study. Percentage of CD38^- HLADR⁺ cells among naive (G), CM (H), and EM (I) CD8⁺ T cells. * Significant difference between the TB IRIS and No IRIS groups. ^# Significant difference between the TB IRIS and Other IRIS groups. ‡ Significant difference between Other IRIS and No IRIS groups. Symbols in brackets denote tendencies (p < 0.1). Two-group differences were determined only when the Kruskal-Wallis test showed overall group effects. Values correspond to each group mean ± 1SEM.

Figure 3

Frequencies and absolute counts of activated CD8 T cell subsets. A) Frequencies of activated CD8 T cell subsets according to CD38 and HLADR expression during follow up. The numbers in each quadrant correspond to the mean of all patients’ percentage of the subpopulation among all CD8 T cells. B) Mean ± 1 SEM absolute counts of CD38⁺ HLADR^- CD8 T cells. C) Mean ± 1 SEM absolute counts of CD38⁺ HLADR⁺ CD8 T cells. D) Mean ± 1 SEM absolute counts of CD38^- HLADR⁺ CD8⁺ T cells. * Significant difference between the TB IRIS and No IRIS groups. # Significant difference between the TB IRIS and Other IRIS groups. Two-group differences were determined only when the Kruskal-Wallis test showed overall group effects.

Figure 4

Expansion of activated naive CD8 T cells during TB IRIS. Zebra plots of CD8⁺ T cell maturation subpopulations according to CCR7 and CD45RA expression (see Methods) in a week 8 sample from a TB IRIS patient (A), and of week 8 sample from an Other IRIS patient who developed CMV retinitis at week 8 (C), showing the group’s mean % naive of CD8⁺ T cells. (B) Zebra plots showing activated subsets of naive CD8 T cells from the TB IRIS patient and the TB IRIS group’s mean % CD38⁺ HLADR⁺ of naive CD8 T cells at week 8. (D) Zebra plot showing activated subsets of naive CD8 T cells from the CMV IRIS patient and the Other IRIS group’s mean % CD38⁺ HLADR⁺ of naive CD8 T cells at week 8. (E) Absolute counts of CD38⁺ HLADR⁺ naive CD8⁺ T cells throughout the study. Values correspond to each group’s mean ± 1 SEM. * Significant difference between the TB IRIS and No IRIS groups # Significant difference between the TB IRIS and Other IRIS groups. Two-group differences were determined only when the Kruskal-Wallis test showed overall group effects.