Long-Term Clinical and Immunological Profile of Kidney Transplant Patients Given Mesenchymal Stromal Cell Immunotherapy

Abstract

We report here the long-term clinical and immunological results of four living-donor kidney transplant patients given autologous bone marrow-derived mesenchymal stromal cells (MSCs) as part of a phase 1 study focused on the safety and feasibility of this cell therapy. According to study protocols implemented over time, based on initial early safety findings, the patients were given MSC at day 7 posttransplant (n = 2) or at day -1 pretransplant (n = 2) and received induction therapy with basiliximab and low-dose rabbit anti-thymocyte globulin (RATG) or RATG alone, and were maintained on low-dose ciclosporin (CsA)/mycophenolate mofetil (MMF). All MSC-treated patients had stable graft function during the 5- to 7-year follow-up, without increased susceptibility to infections or neoplasm. In three MSC recipients, but not historical control patients, circulating memory CD8+ T cell percentages remained lower than basal, coupled with persistent reduction of ex vivo donor-specific cytotoxicity. Two patients showed a long-lasting increase in the regulatory T cell/memory CD8+ T cell ratio, paralleled by high circulating levels of naïve and transitional B cells. In one of these two patients, CsA was successfully discontinued, and currently the low-dose MMF monotherapy is on the tapering phase. The study shows that MSC therapy is safe in the long term and could promote a pro-tolerogenic environment in selected patients. Extensive immunomonitoring of MSC-treated kidney transplant recipients could help selection of patients for safe withdrawal of maintenance immunosuppressive drugs (NCT00752479 and NCT02012153).

Keywords: B cells; kidney transplantation; memory CD8+ T cells; mesenchymal stromal cells; regulatory T cells; tolerance.

Figures

Figure 1

Clinical protocols of MSC-treated kidney transplant recipients implemented at our center. The first two living-related donor kidney transplant recipients adopted the initial clinical protocol (Clinical protocol #1), where autologous bone marrow-derived MSC were infused 7 days after transplantation and induction therapy comprised basiliximab and low-dose RATG. Since in both patients transient acute renal insufficiency occurred due to engraftment syndrome, the protocol was revised (Clinical protocol #2), and two subsequent patients received pretransplant MSC infusion and induction therapy with low-dose RATG alone. Abbreviations: CsA, ciclosporin; MMF, mycophenolate mofetil; MSC, mesenchymal stromal cell; RATG, rabbit anti-thymocyte globulin.

Figure 2

The rate of glomerular filtration rate (GFR) decline in the long term after transplantation. Panel (A) represents median slope of GFR in the first four patients given mesenchymal stromal cells (MSC) along with their respective follow-up (open circles) and in control kidney transplant recipients given either basiliximab/low-dose RATG or low-dose rabbit anti-thymocyte globulin (RATG) alone (closed circles) over 7 and 5 years posttransplantation, respectively. Measured GFR decreased to a significantly lower rate in the first four MSC-treated patients compared with living-related donor kidney transplant recipients in the basiliximab/low-dose RATG control group (n = 3) (−0.278 vs −1.834 mL/min/1.73 m2/year, respectively, P < 0.001). Panel (B) shows individual slopes of measured GFR in the first four patients given MSC along their respective follow-up.

Figure 3

Profile of circulating memory CD8+ T cells and regulatory T cells. Percentages of CD45RO+RA− memory CD8+ T cells (on CD3+CD8+ T cells) in mesenchymal stromal cell (MSC)-treated patients #1 (blue circles), #2 (orange circles) and in control kidney transplant recipients given basiliximab/low-dose rabbit anti-thymocyte globulin (RATG) induction therapy but not MSC (white histograms) during the follow-up (A) and in MSC-treated patients #3 (green circles) and # 4 (violet circles) and in control kidney transplant recipients given induction therapy with low-dose RATG alone but not MSC (gray histograms) during the follow-up (B). Panels (C,D) represent profile of percentages of regulatory T cells (Tregs) (Foxp3+CD127− on CD4+CD25high T cells) during the follow-up in MSC-treated patients #1 and #2 and in control patients given basiliximab/low-dose RATG (white histograms) and in MSC-patients #3 and #4 and in control patients given induction therapy with low-dose RATG alone (gray histograms), respectively. Ratio of Treg/memory CD8+ T cell percentages in MSC-treated patients and in control groups are shown in panels (E,F). Data from controls are mean ± SEM, *P < 0.05 vs pretransplant values.

Figure 4

Profile of naïve and transitional B cell counts in the peripheral blood. Number (A) and percentages [on total CD3−CD19+ B cells (B)] of naïve IgD+CD27− B cells in the peripheral blood during the follow-up in mesenchymal stromal cell (MSC)-treated patients #1 (blue circles) and #2 (orange circles) and in control kidney transplant recipients given basiliximab/low-dose rabbit anti-thymocyte globulin (RATG) induction therapy but not MSC (white histograms) during the follow-up (left panels) and in MSC-treated patients #3 (green circles) and #4 (violet circles) and in control kidney transplant recipients given induction therapy with low-dose RATG alone but not MSC (gray histograms) (right panels). Number (C) and percentages [on total CD3−CD19+ B cells (D)] of CD24highCD38high transitional B cells in MSC-treated patients #1 and #2 and in control patients given basiliximab/low-dose RATG (white histograms) (left panels) and in MSC-patients #3 and #4 and in control patients given induction therapy with low-dose RATG alone (gray histograms) (right panels), during the follow-up. Data from controls are mean ± SEM, P = NS.

Figure 5

IFNγ ELISpot assay against donor or third-party cells. Frequency of IFNγ-producing T cells after overnight incubation with donor (A) or third-party (B) cells of peripheral blood mononuclear cells isolated from mesenchymal stromal cell-treated patient #1 (blue circles), patient #2 (orange circles), patient #3 (green circles), and patient #4 (violet circles) and in control kidney transplant patients [white histograms; n = 6 given basiliximab/low-dose rabbit anti-thymocyte globulin (RATG) and n = 1 given only low-dose RATG induction therapy] during the follow-up. The dotted horizontal line represents the limit of test positivity (25 spots/300.000 responder cells). Data from controls are mean ± SEM; *P < 0.05 vs time 0.

Figure 6

Cell-mediated lympholysis against donor or third-party cells. Percentage of specific lysis of donor (A) or third-party (B) cells after 4 h incubation of 51Cr-labeled donor cells with peripheral blood mononuclear cells isolated from mesenchymal stromal cell-treated patient #1 (blue circles), patient #2 (orange circles), patient #3 (green circles), and patient #4 (violet circles) and in control kidney transplant patients [white histograms; n = 6 given basiliximab/low-dose rabbit anti-thymocyte globulin (RATG) and n = 1 given only low-dose RATG induction therapy] during the follow-up. Data from controls are mean ± SEM.

Figure 7

Profile of circulating NK cell subsets. (A) Profile of NK cell counts in mesenchymal stromal cell (MSC)-treated patients (n = 4), and kidney transplant patients undergoing induction therapy with bas/low dose rabbit anti-thymocyte globulin (RATG) (n = 6) or with low-dose RATG alone (n = 5) from 1 to 5 years posttransplant. Data are mean ± SEM. P = NS. (B) Gating strategy for identification of natural killer cell subsets and natural killer T cells. In the CD45+ live singlet cells, the population of CD3− lymphocytes was gated on the morphology based on SSC and plotted for CD16 and CD56 expression. CD16+CD56dim cells positive for CD11b expression and negative for CD27 were identified as cytotoxic NK cells. The CD56bright CD16−/+ NK cells were divided into CD27+ and CD27− subpopulations. NKT cells were defined as CD3+CD56+ cells. The frequency (% on the gated CD3− population) of CD56dim NK cells, CD56bright NK cells negative or positive for CD27 expression, and of NKT cells (% on CD3+ cells) in patients given MSC infusion and in kidney transplant patients undergoing induction therapy with bas/low dose RATG or with low-dose RATG alone from 1 to 5 years posttransplant are reported in panels (C–F), respectively. Data are mean ± SEM, *P < 0.05 vs MSC-treated patients and control patients given low-dose RATG at the same time point.

Figure 8

Profile of circulating monocytes. (A) Gating strategy for identification of classical, non-classical, and intermediate monocytes in peripheral blood mononuclear cells. In the CD45+ live singlet cells, the population of CD3− leukocytes was gated on the morphology based on SSC and plotted for CD14 expression. Classical monocytes were identified as CD64+ cells on gated CD33+HLADR+ cells in the CD14+CD16− cell population, intermediate monocytes as CD64+ cells on gated CD33+HLADR+ in the CD14+CD16+ cell population, and non-classical monocytes as CD64− cells on gated CD33+HLADR+ cells in the CD14−CD16+ cell population. Percentages of classical monocytes (% on the gated CD3− population) and their HLA-DR expression in patients given mesenchymal stromal cell (MSC) infusion (n = 4) and in kidney transplant patients undergoing induction therapy with bas/low dose rabbit anti-thymocyte globulin (RATG) (n = 6) or with low-dose RATG alone (n = 5) from 1 to 5 years posttransplant are reported in panels (B,C), respectively. Data are mean ± SEM, P = NS. Table in panel (D) reported mean ± SEM of intermediate and non classical monocytes; P = NS.

Figure 9

Histologic analysis of protocol kidney graft biopsy, profile of circulating B and T cell subsets, and immunologic functional assays of memory T cell response in patient #5, who received pretransplant infusion of mesenchymal stromal cell (MSC) in the setting of induction therapy with basiliximab/low-dose rabbit anti-thymocyte globulin (RATG). Patient #5, a 42-year-old woman with end-stage renal disease secondary to tubular interstitial nephropathy, received a kidney transplant from her sister, mismatched for two HLA alleles. She was given MSC before surgery (day −1) and induction therapy with the combination of basiliximab and low-dose RATG. A representative image of Gomori’s trichrome staining on 1-year protocol kidney graft biopsy (original magnification 100×) of MSC-treated patient #5 is shown (A). Percentages of circulating CD45RO+RA− memory CD8+ T cells (on CD3+CD8+ T cells) (B), percentages of circulating regulatory T cells (Tregs) (Foxp3+CD127− on CD4+CD25high T cells) (C), and ratio of the percentages of Tregs to CD45RO+RA− memory CD8+ T cells (D) in patient #5 (yellow circles) and in control kidney transplant recipients given basiliximab/low-dose RATG induction therapy but not MSC (white histograms) from baseline (pretransplant) to 1-year posttransplant. Number of circulating naïve IgD+CD27− B cells (E) and CD24highCD38high transitional B cells (F) in patient #5 (yellow circles) and in control kidney transplant recipients given basiliximab/low-dose RATG induction therapy (white histograms) from baseline to 1-year posttransplant. ELISpot for IFNγ (G) and CD8+ T cell function by T cell-mediated lympholysis [as percentage of specific lysis at 50:1 effector-target ratio (H)] toward donor or third-party antigens in patient #5 (yellow circles) and in control kidney transplant recipients given basiliximab/low-dose RATG induction therapy (white histograms) on peripheral blood mononuclear cells collected pretransplant and at 1 year after transplantation. Data are mean ± SEM.