Modulation of B Cells and Homing Marker on NK Cells Through Extracorporeal Photopheresis in Patients With Steroid-Refractory/Resistant Graft-Vs.-Host Disease Without Hampering Anti-viral/Anti-leukemic Effects

Abstract

Graft-vs.-host disease (GvHD), a severe complication of allogeneic hematopoietic stem cell transplantation, significantly affects the post-transplant morbidity and mortality. Systemic steroids remain the gold standard for the initial management of GvHD. However, up to 60% of patients will not sufficiently respond to steroids. Extracorporeal photopheresis (ECP), a cell-based immunotherapy, has shown good clinical results in such steroid-refractory/resistant GvHD patients. Given its immunomodulatory, but not global immunosuppressive and steroid-sparing capacity, ECP constitutes an attractive option. In the case of GvHD, the balance of immune cells is destroyed: effector cells are not any longer efficiently controlled by regulatory cells. ECP therapy may restore this balance. However, the precise mechanism and the impact of ECP on anti-viral/anti-leukemic function remain unclear. In this study, 839 ECP treatments were performed on patients with acute GvHD (aGvHD) and chronic GvHD (cGvHD). A comprehensive analysis of effector and regulatory cells in patients under ECP therapy included multi-parametric flow cytometry and tetramer staining, LuminexTM-based cytokine, interferon-γ enzyme-linked immunospot, and chromium-51 release assays. Gene profiling of myeloid-derived suppressor cells (MDSCs) was performed by microarray analysis. Immunologically, modulations of effector and regulatory cells as well as proinflammatory cytokines were observed under ECP treatment: (1) GvHD-relevant cell subsets like CD62L+ NK cells and newly defined CD19hiCD20hi B cells were modulated, but (2) quantity and quality of anti-viral/anti-leukemic effector cells were preserved. (3) The development of MDSCs was promoted and switched from an inactivated subset (CD33-CD11b+) to an activated subset (CD33+CD11b+). (4) The frequency of Foxp3+CD4+ regulatory T cells (Tregs) and CD24+CD38hi regulatory B cells was considerably increased in aGvHD patients, and Foxp3+CD8+ Tregs in cGvHD patients. (5) Proinflammatory cytokines like IL-1β, IL-6, IL-8, and TNF-α were significantly reduced. In summary, ECP constitutes an effective immunomodulatory therapy for patients with steroid-refractory/resistant GvHD without impairment of anti-viral/leukemia effects.

Keywords: ECP; GvHD; anti-leukemic effect; anti-viral effect; effector cells; immunomodulation; proinflammatory cytokines; regulatory cells.

Figures

Figure 1

Differentiation and education of NK cell populations by ECP in aGvHD patients. The assessment of CD56briCD16− NK cells (A) and CD56dimCD16+ NK cells (C) before and after ECP therapy shows that ECP treatment can promote the development of NK cells from CD56briCD16− NK cells to CD56dimCD16+ NK cells as well as educate NK cells by decreasing expression of NKG2D and CD62L (B,D). *means p < 0.05.

Figure 2

The role of CD19hiCD20hi B cells in cGvHD. (A) shows representative dot plots of CD19hiCD20hi B cells among HD, aGvHD and cGvHD groups. (B) displays the frequency of CD19hiCD20hi B cells in both aGvHD and cGvHD groups prior to ECP treatment. (C) Characterization of CD19hiCD20hi B cells showed significantly lower expression of BAFF-R and CD38 but slightly increased CD24 expression. (D,E) When compared to CD19+CD20+ B cells, CD19hiCD20hi B cells showed a different component pattern, a significantly higher BAFF-R+CD38− proportion and memory B cells. Dashed lines represent the corresponding median value of frequencies observed in 25 HDs. Differences in cell frequency between different groups were assessed by Independent T test. In all tests, a p-value < 0.05 was considered to be statistically significant. *means p < 0.05.

Figure 3

CD14+HLA-DR−/low MDSC subpopulations in the peripheral blood of GvHD patients with ECP treatment. The immunophenotype of MDSCs was assessed by flow cytometry. (A) Different components of inactivated, transitional and activated subsets were observed within CD14+HLA-DR−/low MDSCs among aGvHD patients, cGvHD patients and healthy donors (HDs) suggesting a development from inactivated into activated MDSCs. (B) A reduction of inactivated MDSCs was observed after ECP therapy in aGvHD patients. (C) The volcano plot shows the gene expression between activated MDSCs and inactivated MDSCs. The horizontal axis represents the fold change in intensity and the vertical axis represents statistical significance (Log Odds). The bar chart indicates the differential gene expression between activated and inactivated MDSCs. Gene expression was assessed after adjustment by the Benjamini-Hochberg procedure. Differences in cell frequency between different groups were assessed by paired-sample T test. In all tests, a p-value < 0.05 was considered to be statistically significant. *means p < 0.05.

Figure 4

Immunomodulation of regulatory T and B cells through ECP. The percentages of CD8+ Tregs (A), FoxP3+CD25+CD4+ Tregs (B), and CD24+CD38hi Bregs (C) were monitored in patients with aGvHD and cGvHD before and after ECP therapy. Foxp3+CD8+ Tregs significantly increased under ECP therapy in both aGvHD and cGvHD patients, along with significant up-regulation of Foxp3+CD4+ Tregs and Bregs in aGvHD patients, as assessed by paired-sample T test. * means p < 0.05.

Figure 5

A fast reduction of proinflammatory cytokines IL-1β (A), IL-6 (B), IL-8 (C), and TNF-α (D) was observed in all patients. Patient #7 showed a rebound of IL-1β and TNF-α after rapid steroid reduction. Eventually the level of both cytokines decreased when ECP was continued. Dashed lines represent the corresponding median value of cytokine levels observed in healthy donors. The frequency of ECP cycles is indicated on the x-axis. The black bars below the x-axis indicate a high frequency of ECP treatment during the first 12 weeks (twice per week) followed by a gray bar representing a reduced frequency (twice every second week) in weeks 13–28.

Figure 6

Impact of ECP therapy on anti-viral and anti-leukemic immune responses. (A) The representative dot plots with the frequency of CMV-specific CD8+ T cells are shown at different time (T) points before (T1) and after (T2 and T3) ECP treatment in aGvHD patient #5 and cGvHD patient #13. (B) The distribution of TCM, TN, TEM, and TE within the CMV specific CD8+ T cells in patients #5 and #13 is indicated. (C) The secretion of IFN-γ by virus specific T cells was measured by IFN-γ ELISpot assay. The bar chart shows the overview of the INF-γ secretion by CD8+ T cells in seven patients under ECP treatment. There is no significant difference among T1, T2, and T3 (p ≥ 0.05, one-way ANOVA test). Under ECP treatment, the frequency of CMV-specific CD8+ T cells was maintained. Most cells were TE cells followed by TEM cells. The function of these cells in terms of IFN-γ release kept stable. (D) The dynamic changes of CD4+CD8+ T cells, γδ T cells and NKT cells in aGvHD (upper panel) and cGvHD patients (lower panel) under the ECP treatment. Cell frequencies were not significantly different between before and under ECP treatment groups, which assessed by Paired sample T test. (E) A 4-h 51Cr release assay was performed to test the NK activity, which was calculated by the following formula: % specific lysis = [c.p.m. (experimental release)–mean c.p.m. (spontaneous release)]/[mean c.p.m. (maximal release)–mean c.p.m. (spontaneous release)] × 100. The box chart shows the NK activity against K562 cells at two different time points in aGvHD group and cGvHD group. There was no significant difference, as assessed by Paired sample T test. Each box represents three independent patients. The NK cell function was stable over the time of ECP treatment. The dashed lines represent the corresponding median value of frequencies observed in 25 healthy donors. In all tests, a p < 0.05 was considered to be statistically significant.

Figure 7

Cell population dynamics are displayed with steroid dosing and clinical parameters in representative patients under ECP therapy.

Figure 8

Schematic overview of mechanisms of immunomodulation in aGvHD patients under ECP therapy. (A) ECP is a cell-based immunotherapy, involving (i) apheresis with ex vivo collection of peripheral mononuclear cells, (ii) photoactivation with exposure of leukocyte-enriched plasma to the photosensitizing agent 8-methoxypsoralen and ultraviolet A light which results in crosslinking of the pyrimidine bases in DNA leading to cell death through apoptosis, (iii) reinfusion of the ECP-treated cells to the patient. (B) Apoptosis of ECP-treated cells play a key role in vivo. Engulfing these apoptotic cells by immature dendritic cells results in a tolerogenic phenotype and promotes tolerance through the secretion of immunosuppressive cytokines such as IL-10 and TGF-β as well. Upregulation of activated MDSCs, Th2, vδ2+ T cells, FoxP3+ Tregs, double negative (DN) T cells and Bregs result in an overall increase in immune tolerance, accompanied by a decrease of immune effector cells like IL-17+ T cells and Th1 cells as well as education of TE/EM cells via decreasing CD62L expression. Besides these, ECP promotes the NK cell differentiation from CD56bri to CD56dim NK cells with loss of expression of NKG2D and CD62L.