Dynamics of T-Lymphocyte Activation Related to Paradoxical Tuberculosis-Associated Immune Reconstitution Inflammatory Syndrome in Persons With Advanced HIV

Abstract

Most persons living with HIV (PLWH) experience a significant restoration of their immunity associated with successful inhibition of viral replication after antiretroviral therapy (ART) initiation. Nevertheless, with the robust quantitative and qualitative restoration of CD4+ T-lymphocytes, a fraction of patients co-infected with tuberculosis develop immune reconstitution inflammatory syndrome (TB-IRIS), a dysregulated inflammatory response that can be associated with significant tissue damage. Several studies underscored the role of adaptive immune cells in IRIS pathogenesis, but to what degree T lymphocyte activation contributes to TB-IRIS development remains largely elusive. Here, we sought to dissect the phenotypic landscape of T lymphocyte activation in PLWH coinfected with TB inititating ART, focusing on characterization of the profiles linked to development of TB-IRIS. We confirmed previous observations demonstrating that TB-IRIS individuals display pronounced CD4+ lymphopenia prior to ART initiation. Additionally, we found an ART-induced increase in T lymphocyte activation, proliferation and cytotoxicity among TB-IRIS patients. Importantly, we demonstrate that TB-IRIS subjects display higher frequencies of cytotoxic CD8+ T lymphocytes which is not affected by ART. Moreover, These patients exhibit higher levels of activated (HLA-DR+) and profilerative (Ki-67+) CD4+ T cells after ART commencenment than their Non-IRIS counterparts. Our network analysis reveal significant negative correlations between Total CD4+ T cells counts and the frequencies of Cytotoxic CD8+ T cells in our study population which could suggest the existance of compensatory mechanisms for Mtb-infected cells elimination in the face of severe CD4+ T cell lymphopenia. We also investigated the correlation between T lymphocyte activation profiles and the abundance of several inflammatory molecules in plasma. We applied unsupervised machine learning techniques to predict and diagnose TB-IRIS before and during ART. Our analyses suggest that CD4+ T cell activation markers are good TB-IRIS predictors, whereas the combination of CD4+ and CD8+ T cells markers are better at diagnosing TB-IRIS patients during IRIS events Overall, our findings contribute to a more refined understanding of immunological mechanisms in TB-IRIS pathogenesis that may assist in new diagnostic tools and more targeted patient management.

Trial registration: ClinicalTrials.gov NCT00933790.

Keywords: IRIS pathogenesis; T cell activation; T lymphocytes; TB-HIV coinfection; inflammation.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2021 Tibúrcio, Barreto-Duarte, Naredren, Queiroz, Anbalagan, Nayak, Ravichandran, Subramani, Antonelli, Satagopan, Anbalagan, Porter, Sher, Swaminathan, Sereti and Andrade.

Figures

Figure 1

Assessment of both CD4+ and CD8+ T cell frequencies in HIV-infected individuals prior to antiretroviral therapy (ART) initiation and during TB-IRIS manifestations. Comparison of CD4+(A) and CD8+(B) T cell frequencies in patients who developed TB-IRIS and those who did not at ART initiation (Baseline) and at the time of TB-IRIS manifestations (Timepoint 1), which occurred between 2 and 6 weeks after ART initiation. (C) CD4/CD8 T cell ratios at the aforementioned conditions. (D) Scatterplot displaying CD4+ and CD8+ T cell frequencies prior to ART initiation. Patients who developed TB-IRIS are depicted by green dots whereas those who did not experience TB-IRIS manifestations are represented by purple dots. Dotted lines on the X and Y axes represent the median value of CD4+ and CD8+ T cell counts within the entire study population, respectively. The shadowed quadrant represents those individuals with CD8+ T cell counts above the median value and CD4+ T cell frequencies below the median value for the entire population. Data were analyzed using the Mann-Whitney test or Wilcoxon matched-pairs test for paired analyses within each study group (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Figure 2

Evaluation of T lymphocyte activation markers in TB-HIV co-infected patients prior to and following ART initiation. Expression of flow cytometry markers related to T cell activation (HLA-DR) (A), exhaustion (PD-1) (B), proliferation (Ki67) (C), and cytotoxicity (Granzyme B) (D) were assessed in whole blood samples from TB-HIV co-infected individuals at baseline (BL) pre-ART and at the time of TB-IRIS manifestation (Timepoint [TP]1), which was 2-6 weeks on ART. Data were analyzed using the Mann-Whitney test or Wilcoxon matched-pairs test for paired analyses within each study group (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Figure 3

ART initiation and TB-IRIS events are associated with coordinated alterations in the relationship between markers of T lymphocyte activation at Baseline (pre-ART) and 2-6 weeks following ART initiation. (A) Network analysis of T cell populations expressing activation-related markers; correlation matrices were developed with bootstrap (100 ×). (B) Networks depicting only significant correlations (p < 0.05), with Spearman rank threshold ± 0.5; graph showing interactome network density among the clinical groups. (C) A heatmap was employed to assess the contribution of each node based on the number of statistically significant correlations; yellow lines (edges) represent positive correlations between nodes (T cell population expressing activation-related markers), whereas blue lines depict negative correlations.

Figure 4

Associations between both CD4+ and CD8+ T lymphocytes expressing activation-related biomarkers and plasma levels of several inflammatory biomarkers. A heatmap was employed to depict the strength of association between CD4+ and CD8+ T cells expressing activation-related biomarkers and the level of circulating cytokines and chemokines at Baseline (prior to ART) and after approximately 2 to 6 weeks following ART initiation (Timepoint 1) for Non-IRIS patients (A) and individuals who experienced TB-IRIS manifestations (B). Statistically significant correlations (P < 0.05) after adjustment for multiple measurements are highlighted with bold squares.

Figure 5

Assessment of CD4+ and CD8+ T lymphocyte activation markers to predict TB-IRIS prior to ART. In order to create predictive models that distinguish TB-IRIS from non-IRIS patients prior to ART, unsupervised machine learning techniques were employed for either CD4+ or CD8+ T cell activation markers. In addition, a model encompassing both CD4+ and CD8+ T cell activation markers was also created. For each model, unsupervised cluster analyses (Ward’s method) and principal component analyses were employed. In order to estimate the accuracy of these models in discriminating patients, a receiver-operating characteristic curve analysis was employed.

Figure 6

Combination of T cell activation markers to diagnose IRIS after ART initiation. In order to create diagnostic models that distinguish TB-IRIS from non-IRIS patients after ART initiation, unsupervised machine learning techniques were employed for either CD4+ or CD8+ T cell activation markers. In addition, a model encompassing both CD4+ and CD8+ T cell activation markers was also created. For each model, unsupervised cluster analyses (Ward’s method) and principal component analyses were employed. In order to estimate the accuracy of these models in discriminating patients, a receiver-operating characteristic curve analysis was employed.