Evaluating Activated Regulatory T Cells as a Biomarker of Chronic Allograft Inflammation in Pediatric Kidney Transplant Recipients

Abstract

Background: There is a need for noninvasive immunological biomarkers that can identify stable kidney allograft immune quiescence to inform individualized immunosuppression.

Methods: We conducted a cross-sectional, pilot cohort study evaluating the relative abundance of regulatory T cells (Tregs) to effector T-cell (Teff) populations as a surrogate marker of long-term graft tolerance. We obtained fresh peripheral blood mononuclear cell samples from stable pediatric kidney transplant recipients, most with recent surveillance biopsies to identify the presence or absence of chronic inflammation. Tregs were sub-phenotyped as naïve, memory, and activated Tregs (aTreg). Treg/Teff ratios were modeled for association with chronic inflammation and in the context of potential clinical features.

Results: Twenty-seven patient samples were included on standard immunosuppression (tacrolimus, mycophenolate, and prednisone) with a mean age of 9.2 ± 5.0 years, at 30.2 ± 21.7 months posttransplant. The ratio of aTreg (FOXP3⁺⁺CD45RA^-) to Th17 cells (CD4⁺IL-17⁺) was significantly greater in patients without inflammation than in patients with graft inflammation (p < 0.01). Similarly, there was a trend toward greater aTreg/CD4⁺ T cells and aTreg/CD8⁺ Teff in patients without inflammation (p = 0.05 and 0.09, respectively). There was no significant association for inflammation with naïve or memory Treg/Teff ratios. Multiple logistic regression with all three aTreg/Teff ratios modeled allograft inflammation with high sensitivity and specificity (AUC = 0.83, 95% CI 0.67-0.98).

Conclusions: The proportion of peripheral blood aTregs/Teff cells in this pilot cohort of stable pediatric kidney transplant recipients was associated with immune quiescence. These data support further investigation into aTreg/Teff monitoring to inform precision immunosuppressive treatment.

Keywords: Treg/Teff; immune biomarkers; kidney transplantation; regulatory T cells.

FIGURE 1

Flow of study participants. AR, acute rejection; BCCH, British Columbia Children's Hospital.

FIGURE 2

T cells were characterized as live CD45⁺CD3⁺ and further gated as CD8⁺CD4⁻ (CD8⁺ T cells), CD4⁺CD8⁻ (CD4⁺ T cells), CD4⁺IL‐17⁺ (Th17 cells), and CD4⁺CD25^highCD127^lowFOXP3⁺ (total Tregs). Total Tregs were subdivided into CD45RA⁻ (memory Tregs), CD45RA⁺ (naïve Tregs), and FOXP3⁺⁺CD45RA⁻ (activated Tregs). (A) Frequency of Treg populations in patients with no inflammation and with inflammation. (B) Ratio of Treg populations to CD8⁺ T cells and (C) Th17 cells. Differences between two groups were assessed with a Mann–Whitney U test. The plots display the median, interquartile range, and minimum and maximum values. Inflam; inflammation.

FIGURE 3

Modeling pediatric kidney graft inflammation with serum Treg levels. (A–C) Multivariate logistic regression using a combination of activated Treg (% of CD4⁺), activated Treg/CD8⁺ T cells, and activated Treg/Th17 cells to model graft inflammation. (D) Combining all three activated Treg‐to‐effector T cell ratios to model graft inflammation in a multivariate logistic regression. aTreg, activated Treg; AUC, area under the ROC curve; CI, confidence interval.

FIGURE 4

Combining clinical and cellular patient predictors to best model graft inflammation. (A) Multivariate logistic regression using a combination of clinical variables and the activated Treg/CD4⁺ T cells, Treg/CD8⁺ T cells, and Treg/Th17 ratios to model graft inflammation. AUC, area under the ROC curve; CI, confidence interval.

FIGURE 5

Proportion of activated Tregs is not correlated to Tacrolimus or MMF exposure. Pearson's correlation between the ratio of activated Treg/CD4⁺ T cells, activated Treg/CD8⁺ T cells, or activated Treg/Th17 cells and (A) tacrolimus (TAC) exposure or (B) mycophenolate mofetil (MMF) dosage in pediatric renal transplant recipients (n = 25).