T cell immuno-phenotyping : a source of predictive biomarkers for autoimmune hepatitis relapse

Abstract

Relapse after immunosuppression (IS) treatment withdrawal is frequent in patients with Autoimmune Hepatitis (AIH), and non-invasive biomarkers predictive of this risk are lacking. We assessed the frequency of circulating T cell subsets as potential biomarkers of disease activity and predictor of the risk of relapse after IS withdrawal. Serum levels of the cytokine B-cell Activating Factor (BAFF) were also investigated. Blood samples from 58 patients with active AIH, 56 AIH patients in remission, and 31 patients with NASH were analyzed. The frequency of activated CD4+ T peripheral helper (TPH) cells (CD4+CD45RA-CXCR5-PD1+CD38+) and of activated CD8+ T cells (CD8+CD45RA-PD1+CD38+) were assessed by flow cytometry. BAFF levels were determined by ELISA. Activated TPH and CD8+ T cell frequencies were significantly increased in patients with active AIH compared to remission AIH or NASH (TPH: 0.88% of total CD3+ vs. 0.42% and 0.39% respectively, p < 0.0001; CD8+ subset: 1.42% vs. 0.09% and 0.11% p < 0.0001). Among patients in remission undergoing treatment withdrawal (n = 18), those with increased frequencies of activated TPH (> 0.5% of total CD3+) and/or activated CD8+ T cells (> 0.18% total CD3+) had a higher risk of relapse (80% vs. 15% after 2 years, p = 0.0071). High BAFF serum concentration (> 213pg/ml) was also associated to a higher risk of relapse (57% vs. 11%, p = 0.0452). In conclusion, high frequency of activated TPH and of activated CD8+, as well as high levels of BAFF, before IS discontinuation, were significantly associated to a greater risk of relapse during the first two years. Thus, they represent promising biomarkers to provide personalized clinical follow-up for patients with AIH.

Keywords: BAFF; Biomarkers; CD38; Peripheral helper T cells; Personalized medicine.

Figure 1

Gating Strategy of total PBMC, to identify and quantify the populations of interest : activated CD8 T cells (CD8+CD45RA-PD1+CD38+) activated TFH cells (CD4+CD45RA-CXCR5+PD1+CD38+); activated TPH cells (CD4+CD45RA-CXCR5-PD1+CD38+). A simple gating strategy #1 consists in gating on memory (CD45RA-) CD3+ T cells, then on CD38+PD1+ T cells to maximize the visibility (percentage between 35 to 0.2%). Isotype controls are used to define the PD1+CD38+ T cells. The fraction of CD4+ and CD8+ T cells is determined on this PD1+CD38+ subset. In the CD4+ fraction, we also isolate T cells based on the CXCR5 expression. A more classical strategy (#2) is shown, which consists in consecutive gating of CD3, then CD4 /CD8, then CXCR5/CD45RA, and finally PD1/CD38. (B) Comparisons of the subset frequencies obtained with two strategies using Spearman correlation test.

Figure 2

Representative dot plots and graph of the frequency of cell subsets. (A) activated TFH cells (CD4+CD45RA-CXCR5+PD1+CD38+); (B) activated TPH cells (CD4+CD45RA-CXCR5-PD1+CD38+) and (C) activated CD8 T cells, in patients with active AIH (AIHa, n = 58), remission AIH (AIHr, n = 56) and NASH (n = 31), in % of total CD3+ T cells. (D) Spearman correlation analysis between the frequencies of activated CD8 T cells and activated TPH cells in patients with active AIH (n = 58). Activated TPH (E) and CD8 T cells (F) in AIH patients in remission, depending on their treatment (Steroids +/- IS n = 12, IS only n = 37, or no treatment n = 7). Kruskal Wallis test and Dunn’s multiple comparisons test were used for A, B, C, E and F.

Figure 3

Comparison of frequency of activated T cell subsets between active AIH (AIHa) and NASH, and between AIHa and AIHr (brackets) by AUROC Analysis for determination of significant cut-off values (Tresh). Frequencies are expressed as percent of total CD3+ T cells.

Figure 4

Graph representation of the frequency of activated TPH cells (A) and activated CD8 T cells (B), in the cohort of AIH patients in remission before treatment withdrawal (n = 18). Dotted lines represent the thresholds determined in Fig. 3. (C - E) Relapse-free survival stratified by the frequency of activated TPH (group high TPH, n = 3; group low TPH, n = 15) (C) or of activated CD8 T cells (group high CD8, n = 4; group low CD8, n = 14) (D), or combined frequencies of activated TPH and/or of CD8 T cells group high TPH/CD8, n = 5; group low TPH/CD8, n = 13) (E), determined by flow cytometry. Log-rank (Mantel Cox) analysis.

Figure 5

BAFF levels and risk of relapse. (A) Quantification of BAFF levels in NASH patients (n = 30) and active AIH (AIHa, n = 55), AIH in remission (AIHr, n = 52). (B) ROC analysis of BAFF levels for AIHa vs. NASH (graph), and AIHa vs. AIHr (brackets). (C) Quantification of the level of BAFF in AIH patients in remission before treatment withdrawal (n = 16), a dotted line represents the threshold determined in B. (D) Relapse-free survival, stratified by the serum level of BAFF measured in AIH patients in remission before treatment withdrawal (group high BAFF, n = 7; group low BAFF, n = 9).

Figure 6

Flow cytometry analysis on fresh whole blood and isolated PBMC. (A) Gating strategy of whole blood sample. (B) gating strategy on freshly isolated PBMC from the same AIHa patient.