Patterns and Cofactors of Polyfunctional Mycobacteria-Specific T-Cell Response Restoration Following 6-Month Antiretroviral Treatment in Children With HIV

Abstract

Background

Despite immune restoration after initiation of antiretroviral treatment (ART), the risk of tuberculosis (TB) persists in children with HIV (CHIV). We determined patterns of immune restoration of mycobacteria-specific T cells following ART in CHIV.

Methods

CD4 and CD8 T-cell activation and memory phenotype and functional profiles before and 6 months after ART were evaluated in peripheral blood mononuclear cells from CHIV enrolled in the PUSH study (NCT02063880) in Nairobi, Kenya. T-cell expression of cytokines and activation-induced markers were measured following stimulation of peripheral blood mononuclear cells with a pool of 300 peptides from TB (MTB300) or staphylococcal enterotoxin B.

Results

Among 47 CHIV (median age, 1.5 years), staphylococcal enterotoxin B–induced Th1 cytokine⁺ and activation-induced marker⁺ CD4 cell frequencies increased significantly after 6 months of ART. Although MTB300-specific CD4 and CD8 cell frequency did not increase after ART, polyfunctional capacity of MTB300-specific CD4 cells expressing combinations of Th1 cytokines with CD40L increased significantly after ART. Baseline age, immune activation, and effector memory CD4 levels were associated with less restoration of MTB300-specific polyfunctional CD4 cells, whereas CD4 percentage and levels of naive CD4 cells following ART were associated with improved MTB300-specific polyfunctional capacity.

Conclusions

Despite increases in Th1 cytokine production, deficits in mycobacteria-specific CD4 cells persisted 6 months after ART, with higher deficits in older CHIV with more immunosuppression, higher immune activation, and lower proportion of naive CD4 cells. These findings may explain persistent TB risk during early ART among CHIV and identify those at highest risk.

Children with HIV (CHIV) are often diagnosed late in the course of their HIV infection and are at increased risk for tuberculosis (TB) disease [1–3]. Mortality is high in CHIV with advanced HIV, particularly in the context of TB coinfection [4, 5]. Antiretroviral treatment (ART) substantially decreases morbidity and mortality in CHIV; however, risk for TB and mortality persists in the context of late HIV diagnosis.

Immune reconstitution following ART differs between children and adults with HIV and between younger and older CHIV. Among children with perinatal HIV in Zimbabwe who initiated ART late (6–15 years old), younger children had an accelerated CD4 T-cell increase [6]. Early ART (initiated <6 months of age) resulted in higher polyfunctional HIV-specific CD4 and CD8 T-cell immune reconstitution as compared with children who started ART after 1 year of age [7]. Age-related differences in immune reconstitution between infants and older CHIV following ART initiation are likely due to shorter duration of HIV infection in infants but could additionally reflect age-related immune differences. CD4 T-cell reconstitution occurs rapidly after ART in adults and children; however, most new CD4 T cells in children are CD45RA⁺ naive cells, in contrast to adults who expand CD45RO⁺ memory cells [8]. Among infants who start ART before 3 months of age, the naive T-cell pool and thymic production of naive cells are key determinants of CD4 T-cell reconstitution [9]. Longer-term studies in CHIV in Zimbabwe and Uganda observed that immune reconstitution was driven by CD45RA⁺ CD31⁺ CD4 T cells, with a decline in activated and proliferating T cells and most inflammatory biomarkers following 3.5 years of ART [10].

In the absence of ART, infants with HIV have substantially lower T-cell responses to Bacille Calmette-Guerin (BCG) vaccine than infants without HIV [11]. However, among older children who have perinatally acquired HIV, survived past infancy, and are ART naive, T-cell responses to mycobacteria may be similar to those of children without HIV, perhaps because these children had close-to-normal induction of immune responses to BCG vaccination at birth, prior to experiencing significant HIV disease progression and immune suppression [12]. Although antimycobacterial immune responses are reconstituted with ART, residual deficits were identified in youth with HIV after ≥1 year of ART [13, 14]. Among CHIV with a median age of 7.6 years in South Africa and no history of TB disease, purified protein derivative (PPD)–specific polyfunctional CD4 T-cell responses were lower in CHIV than children without HIV, with some increase in TNF-α responses but no change in polyfunctional PPD-specific T-cell responses after 1 year of ART [15].

We compared the frequencies and functional capacity of mycobacteria-specific CD4 T-cell responses among hospitalized CHIV who were severely ill, ART naive, and participating in the PUSH trial (Pediatric Urgent Start of Highly Active ART; NCT02063880), before and after 6 months of ART.

METHODS

Participants and Samples

This study used specimens collected in a previously conducted randomized trial [16]. The PUSH study enrolled hospitalized CHIV who were ART naive, aged 0 to 12 years, and randomized to urgent (within 48 hours) vs poststabilization ART. We tested a subset of 47 children in the PUSH study with paired specimens from enrollment (pre-ART) and 6 months after ART. All children were evaluated for TB at enrollment, and cases were classified as confirmed, unconfirmed, or unlikely TB [17].

Antigens

Peptide pools were used at a final concentration of 1 μg/mL: MTB300, CFP-10/ESAT-6, and human cytomegalovirus (CMV) pp65. The CMV pp65 peptide pool was obtained through the AIDS Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health. The CFP-10 (NR-50712) and ESAT-6 (NR-50711) peptide pools were obtained through BEI Resources, National Institute of Allergy and Infectious Diseases, National Institutes of Health. The MTB300 peptide pool consists of 300 peptides derived from Mycobacterium tuberculosis (Mtb) [18]. Staphylococcal enterotoxin B (SEB; Toxin Technology, Inc) was used as a positive control at 1 μg/mL.

Antibodies

The following monoclonal antibodies were used in flow cytometry assays to evaluate T-cell phenotype and functional capacity: anti-CD3, anti-CD4, anti-CD8, anti-CCR7, anti-CD45RA, anti-HLA-DR, anti-CD38, anti-IFN-γ, anti-IL-2, anti-TNF-α, anti-MIP-1β, anti-IL-17A, anti-CD69, and anti-CD40L. Details of antibody fluorochrome conjugations, clones, and manufacturer are provided in Supplementary Table 1.

T-Cell Phenotype, Intracellular Cytokine Staining, and Flow Cytometry

Cryopreserved peripheral blood mononuclear cells were thawed and evaluated by flow cytometry for measurement of CD4 and CD8 T-cell memory and activation phenotype and frequencies of mycobacteria-specific T cells (Supplementary Methods 1A).

Data Analysis and Statistics

Flow cytometry data were analyzed with FlowJo version 10 (BD; Supplementary Methods 1B). Absolute counts of CD4 T cells expressing all combinations of IFN-γ, IL-2, TNF-α, MIP-1β, IL-17A, and CD40L and CD8 T cells expressing all combinations of IFN-γ, IL-2, TNF-α, MIP-1β, and CD107a were exported from FlowJo and analyzed by the COMPASS algorithm (combinatorial polyfunctionality analysis of antigen-specific T-cell subsets) as described previously [19]. COMPASS uses a bayesian computational platform to evaluate antigen-specific T-cell responses across all observable T-cell subsets and was used to generate a functionality score (FS) and polyfunctionality score (PFS) for each antigen-specific CD4 and CD8 T-cell response at each time point as previously described [20].

Statistical analysis was conducted with Prism version 10.2.2 (GraphPad) and Stata version 16 (StataCorp). Differences between paired samples before and after ART were evaluated with a Wilcoxon matched-pairs signed rank test. Linear regression was used to determine cofactors of MTB300-specific CD4 PFS. P values <.05 were considered significant.

Ethics

Caregivers of enrolled children provided written informed consent, and the study was approved by the Kenyatta National Hospital and University of Nairobi Ethical Review Committee, the University of Washington Institutional Review Board, and the Emory University Institutional Review Board.

RESULTS

Characteristics of the Study Population

Among 47 CHIV with paired specimens from pre-ART and 6 months after ART, the median age was 1.5 years (range, 0–10; IQR, 0.7–5.4). Twenty children had a diagnosis of TB disease at enrollment (42.6%; confirmed, n = 2; unconfirmed TB, n = 18). Participants had a high median HIV viral load at baseline (5.79 log₁₀ copies/mL; IQR, 5.08–6.28). Median absolute CD4 count was 713 cells/μL (IQR, 332–1469), and CD4 percentage (CD4%) was 13% (IQR, 9–18). A substantial proportion (81%) was severely immunosuppressed (Table 1).

Immune Restoration in CHIV 6 Months After Initiation of ART

After 6 months of ART, CD4 count and CD4% increased significantly to 1321 cells/μL and 21%. Concurrently, HIV RNA declined by >3-log to 2.29 log₁₀ copies/mL. CD4 and CD8 T-cell activation (CD38⁺ HLA-DR⁺ cells) decreased significantly by 2- to 3-fold (Table 1). In addition, there was an increase in the proportion of naive CD8 T cells from 13.9% to 22.2%, while the proportion of central memory (CM) and effector memory (EM) CD8 T cells decreased from 3.5% to 2.9% and from 32.8% to 25.0%, respectively. By contrast, the CD4 T-cell naive and EM subset distribution did not change significantly over the 6-month period, while CM CD4 T cells increased from 15.3% to 17.6%.

Th1 Cytokine Production Capacity Increases in CHIV After 6 Months of ART

CD4 and CD8 T-cell functional capacity in CHIV after ART was evaluated by stimulation of peripheral blood mononuclear cells with SEB and measurement of IFN-γ, IL-2, TNF-α, IL-17, MIP-1β, activation-induced markers (AIMs; CD69⁺ CD40L⁺), and CD107a by flow cytometry (Figure 1A). The capacity of CD4 cells to produce Th1 cytokines (IFN-γ, IL-2, TNF-α) and coexpress AIMs increased significantly after 6 months of ART, although CD4 cell capacity to produce MIP-1β or IL-17 did not appreciably change following ART (Figure 1B). The capacity of CD8 cells to produce IL-2 and TNF-α increased after ART.

Frequencies of Mycobacteria-Specific CD4 and CD8 T Cells After 6 Months of ART

Mycobacteria-specific CD4 and CD8 T-cell functional capacity was evaluated by stimulation with the MTB300 peptide pool, containing T-cell epitopes from Mtb and BCG. After 6 months of ART, there were decreased frequencies of MTB300-specific MIP-1β⁺ CD4 cells and a trend for increased frequencies of TNF-α⁺ CD4 T cells (Figure 1C). Relative to other cytokine-producing CD8 cells, the highest frequencies of MTB300-specific CD8 cells were CD107a⁺ cytotoxic cells. There was no evidence of restoration of MTB300-specific CD8 cells producing IFN-γ, IL2, or TNF-α. Similar to CD4 cells, MTB300-specific MIP-1β⁺ CD8 cells decreased after 6 months of ART.

Initiation of ART Was Associated With Increased Functional Capacity of Mycobacteria-Specific CD4 T Cells but Not CD8 T Cells

We next used Boolean gating to measure the capacity of cells to produce multiple combinations of effector molecules. MTB300-specific CD4 and CD8 T-cell responses were analyzed by a bayesian computational framework in COMPASS to compute the FS and PFS for each child before and after ART. The FS summarizes the breadth of the effector molecule repertoire while taking into account the magnitude of responses, whereas the PFS summarizes the diversity of the effector molecule repertoire while taking into account the number of effector molecules within a subset. Consistent with increased frequencies of Th1 cytokine–producing CD4 and CD8 cells after ART (Figure 1), total CD4 and CD8 cell FS and PFS to SEB increased significantly after 6 months of ART (Figure 2A, Supplementary Figure 2). MTB300-specific CD4 cell FS and PFS increased significantly after 6 months of ART, although MTB300-specific CD8 cell FS and PFS did not change appreciably following 6 months of ART (Figure 2B).

We next analyzed COMPASS-generated heat maps indicating the posterior probability of a response to further evaluate frequencies of individual subsets of MTB300-specific CD4 and CD8 cells before and after ART (Figure 2C). The frequency of MTB300-specific CD4 cells producing only IL-2 decreased after ART, whereas polyfunctional subsets of IL-2⁺ TNF-α⁺ CD40L⁺, IFN-γ⁺ TNF-α⁺ CD40L⁺, and IFN-γ⁺ IL-2⁺ TNF-α⁺ CD40L⁺ CD4 cells increased after ART (Figure 2D). The proportion of MTB300-specific CD4 T cells producing TNF-α increased after ART (P = .0515; Figure 2E). Frequencies of individual subsets of MTB300-specific CD8 cells did not change following ART (data not shown).

Given that 43% of children in this study were diagnosed with TB disease at enrollment, we next evaluated whether the FS and PFS of MTB300-specific T cells differed according to TB disease classification. The FS and PFS of MTB300-specific CD4 cells increased significantly in children who were unlikely to have TB at the time of ART initiation, whereas the FS and PFS of MTB300-specific CD4 cells in children with TB did not change within the first 6 months of ART (Figure 3A). The FS of MTB300-specific CD8 cells after ART did not differ by TB status (Figure 3B). We also evaluated T-cell responses to CFP-10/ESAT-6 peptide pools, which were overall lower than responses to MTB300 and did not change significantly after ART (Supplementary Figure 3A and 3B). However, among children with TB, a modest increase in PFS of CFP10/ESAT6-specific CD4 T cells was observed after 6 months of ART (P = .0302), with no changes in the FS or PFS in children who were unlikely to have TB at the time of ART initiation (Supplementary Figure 3C). Similar to MTB300, CFP10/ESAT-6–specific CD8cell FS and PFS did not increase after 6 months of ART, regardless of TB status at the time of ART initiation (Supplementary Figure 3B and 3C).

To evaluate whether the increase in MTB300-specific CD4-cell FS and PFS after ART was unique to mycobacteria-specific cells, we evaluated CMV-specific T cells. There was a modest increase in PFS and a trend of increased FS of CMV-specific CD4 and CD8 cells after 6 months of ART (Supplementary Figure 4). The greater increase in MTB300-specific CD4 FS and PFS as compared with CMV-specific CD4 responses in the same children suggests that the mycobacteria-specific T-cell response may be more amenable to functional restoration early after ART than T cells specific for other pathogens.

Cofactors of Increased MTB300-Specific CD4 Polyfunctional Responses Following ART

Baseline levels of MTB300-specific CD4 polyfunctional capacity were associated with baseline CD4% and inversely associated with baseline age. We next evaluated cofactors of restoration of MTB300-specific CD4 polyfunctional capacity and identified several baseline factors that predicted levels of MTB300-specific CD4 PFS following 6 months of ART (Table 2). Higher baseline age at ART initiation, levels of immune activation (percentage CD38⁺ HLA-DR⁺ CD4 T cells), and CD45RA^– CCR7^– (EM) CD4 cells were all associated with lower MTB300-specfic CD4 PFS at 6 months after ART. Baseline levels of CD45RA⁺ CCR7⁺ (naive) CD4 cells were associated with an increased 6-month CD4 MTB300 PFS. Higher levels of CD4 MTB300 PFS at baseline were associated with higher levels at 6 months post-ART. Several factors measured at 6 months after ART were also associated with CD4 MTB300 PFS. Children with higher CD4% and higher CD45RA⁺ CCR7⁺ (naive) CD4 T cells following 6 months ART had higher MTB300-specific CD4 PFS. Children with higher proportions of CD45RA^– CCR7^– (EM) and CD45RA^– CCR7⁺ (CM) CD4 cells at 6 months after ART had a lower CD4 MTB300 PFS. In stratified analyses of children with TB (confirmed and unconfirmed) and unlikely TB, children with TB had some differences in immunologic cofactors identified than the overall group. The group of children with unlikely TB had similar cofactors as the overall group (Supplementary Table 2).

DISCUSSION

In this study of CHIV who were severely immunosuppressed, we observed changes in several immune parameters following initiation of ART. Over the 6-month period, with expected declines in viral load and increased CD4 counts and percentages, T-cell immune activation decreased substantially; CD4 naive and memory T-cell subsets remained relatively stable, whereas the proportion of naive CD8 T cells increased. The overall Th1 cytokine production capacity of CD4 and CD8 T cells increased after 6 months of ART. Polyfunctional cytokine production capacity of mycobacteria-specific CD4 T cells, but not CD8 T cells, increased significantly after ART.

Immune restoration differs between children and adults with more of an increase in naive T cells in children than adults due to thymic repopulation in children [21–23]. Children in these prior studies ranged from <1 year of age to 15 years, similar to the age range of <1 to 10 years in our cohort. Immune restoration following ART varies by duration of untreated HIV infection, with initiation of ART during acute HIV resulting in more rapid and complete immune restoration than ART initiated during chronic HIV infection [24]. Immune restoration continues throughout the period on ART; thus, duration of ART and duration of untreated HIV influence immune restoration in opposite directions. We assume that our cohort exclusively includes children with vertical transmission, given their age of ≤10 years. Children may have acquired HIV in utero, intrapartum, or postnatally (usually not later than 2 years given typical breastfeeding practices). Thus, we do not expect that children in the cohort had acute HIV at baseline, given that they presented at hospitalization at a median age of 1.5 years. Evaluation of immune restoration in this cohort followed a standardized ART duration (6 months); however, since children ranged from <1 to 10 years of age, there was varied duration of HIV infection pretreatment. We found a small but significant increase in CM CD4 cells and no change in the proportion of naive or EM CD4 cells following ART. These findings differ from 2 prior studies of CHIV in sub-Saharan Africa, which evaluated ART-mediated immune restoration in slightly older CHIV (median age, 5–7 years) and reported increased frequencies of naive CD4 cells after ART [10, 15]. Lack of increased naive CD4 cells in our study may be due to persistent antigenic stimulation in this hospitalized cohort, inclusion of older children without an active thymus, or insufficient follow-up time. In contrast to the relatively stable distribution of memory and naive subsets of CD4 cells in our cohort, the proportion of naive CD8 cells increased substantially and significantly, while CM and EM CD8 decreased significantly, with more decline in the EM population. Our findings of increased naive CD8 cells in young CHIV following ART are consistent with prior studies [8–10, 15].

Prior studies of immune reconstitution of antigen-specific T-cell responses in CHIV undergoing ART vary due to differences in participant age, ART regimens, CD4 levels, nutritional status, and assay/technical differences in measurements of antigen-specific T-cell responses. We evaluated immune responses to BCG- and Mtb-specific antigens, with the premise that most children would reconstitute BCG-primed responses and that fewer would have had prior Mtb infection. ART-mediated restoration of overall T-cell functional capacity was evident by significantly increased frequencies of CD4 cells producing IFN-γ, IL-2, and TNF-α and CD8 cells producing IL-2 and TNF-α following stimulation with SEB. Stimulation with MTB300 did not elicit similar increases in total frequencies of cytokine⁺ cells, which is consistent with a study of CHIV in South Africa in which total frequencies of PPD-specific CD4 T cells producing IFN-γ, IL-2, and TNF-α did not increase significantly in CHIV after 1 year of ART [15]. However, using the COMPASS computational platform, we found evidence of changes in the functional quality of CD4 and CD8 cells after ART, as evidenced by significant increases in the FS and PFS of CD4 and CD8 cells to SEB and in the FS and PFS of CD4, but not CD8, cells following stimulation with MTB300. Moreover, we found evidence of increased proportions of MTB300-specific CD4 cells producing TNF-α after 6 months of ART, consistent with a previous report of increased proportions of TNF-α⁺ PPD-specific CD4 responses in CHIV after 1 year of ART [15]. TNF-α plays an important role in controlling Mtb, as evidenced by reactivation of TB disease in settings where TNF-α signaling is either absent or inhibited [25–27]; thus, increased TNF-α production capacity by TB-specific CD4 cells after ART may be beneficial to enhance immune control of Mtb in CHIV. MTB300-specific CD4 PFS increased more than PFS of CMV-specific T cells, thus suggesting improved early restoration of TB vs CMV responses, albeit much less vigorously than restoration of responses to SEB. The variability of reconstitution of antigen-specific CD4 and CD8 responses to Mtb and CMV following ART was noted in a small study of adults, including those with latent and active TB [28]. Responses to MTB300 include responses to BCG, Mtb, or both and are complex to evaluate in children who have active TB, as active TB could increase or decrease responses independent of reconstitution of BCG responses.

We evaluated cofactors of MTB300-specific CD4 PFS following 6 months of ART and identified several interesting baseline cofactors, including age and immune activation, which were negatively associated with MTB300-specific CD4 PFS after 6 months ART. Older age and immune activation reflect longer duration of HIV with attendant immune compromise and lower baseline MTB300-specific CD4 responses prior to reconstitution. Indeed, we found that baseline MTB300-specific CD4 responses were associated with baseline CD4% and inversely associated with age and that baseline levels of MTB300-specific CD4 PFS correlated with levels at 6 months after ART. Children with higher CD4% and levels of naive CD4 cells at baseline and 6 months post-ART had better restoration of MTB300-specific CD4 PFS. Conversely, children with higher baseline or 6-month post-ART levels of EM CD4 cells or 6-month post-ART CM CD4 cells had lower restoration of MTB300-specific CD4 PFS. The proportion of naive cells may reflect a shorter duration of immune damage from HIV and more ability to reconstitute immune responses in contrast to higher proportion of EM cells. Cofactors of MTB300 PFS in CD4 in the unlikely TB group were similar to overall cofactors, suggesting that these cofactors may predict patterns and cofactors of reconstitution of mycobacteria-specific responses. Slight differences in cofactors identified in the confirmed/unconfirmed TB group likely reflect confounding by immune responses to concomitant Mtb infection and active TB disease.

Strengths of this study include comprehensive analysis of multiple T-cell functions, including Th1 and Th17 cytokine production, AIM, and cytotoxicity, and evaluation of phenotypic and functional profiles of CD4 and CD8 T-cell subsets. Prior studies of ART-mediated restoration of mycobacteria-specific T cells excluded analysis of children with a history of active TB [15]. A substantial portion of CHIV in our cohort were diagnosed with TB at enrollment, thus offering a unique opportunity to further evaluate ART-mediated restoration of mycobacteria-specific T cells stratified by TB classification. This evaluation revealed enhanced reconstitution of mycobacteria-specific CD4 T cells in CHIV who were unlikely to have TB as compared with those with TB.

Limitations of our study included the duration of follow-up of 6 months post-ART in this cohort of CHIV who were severely immunocompromised, although longer-term follow-up is necessary to fully capture the sustained impact of ART on immune restoration and T-cell phenotypic and functional profiles. Our cohort included a wide range of ages (0–10 years), with a low median age (1.5 years) spanning a range of immune development. We did not have a comparator cohort of age-matched children without HIV and did not measure reconstitution of HIV-specific T-cell responses in our cohort. Larger cohorts with more age stratification would be useful to understand age-specific immune restoration patterns. All children were hospitalized at the time of enrollment, with a range of clinical diagnoses. Acute illness of varying etiology could drive the heterogeneity of immune response profiles at the baseline time point.

In summary, future studies require longer-term follow-up of CHIV to better define the capacity of ART to mediate restoration of mycobacteria-specific CD4 and CD8 T-cell responses that are associated with protection against acquisition of Mtb infection and/or progression to TB disease.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Acknowledgments. We thank the children and their caregivers for participation in the study.

Author contributions. Concept and study design: C. L. D., I. N. N., L. M. C., S. M. L., C. M., H. M. O., E. M.-O., D. C. W., G. C. J.-S. Data collection, experimental optimization, and acquisition: C. L. D., I. N. N., L. M. C., W. E. W., R. A. P., J. N. E., L. E. S., C. M., H. M. O., E. M.-O., D. C. W., G. C. J.-S. Provision of study materials and reagents: C. S. L. A., A. S. Analysis and interpretation of data: C. L. D., I. N. N., L. M. C., R. A. P., L. E. S., S. M. L., C. M., H. M. O., E. M.-O., D. C. W., G. C. J.-S. Writing the original draft of the manuscript: C. L. D., G. C. J.-S. All authors contributed to reviewing and editing the final version of the manuscript.

Financial support. This work was supported by grants from the National Institute of Allergy and Infectious Diseases (National Institutes of Health; R01AI142647 to C. L. D. and G. C. J.-S.; P30AI168386 to C. L. D.) and a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (National Institutes of Health; R01HD023412 to G. C. J.-S.).

References

Author notes