Control of graft-versus-host disease with maintenance of the graft-versus-leukemia effect in a murine allogeneic transplant model using retrovirally transduced murine suicidal lymphocytes

Abstract

Objective: Limited clinical trials have validated the hypothesis of controlling graft-versus-host disease (GVHD) arising from stem cell transplant utilizing suicidal T-lymphocytes that have been transduced to express the HSV-TK gene. However, clinical utility has been limited by diminished T-cell function arising from the production process. To evaluate strategies for harnessing the graft-versus-leukemia (GVL) effect while improving the safety and function of suicidal lymphocytes, we have developed techniques to produce fully functional, retrovirally transduced, HSV-TK-positive murine T cells (TK+TC).

Methods: Utilizing a murine major histocompatibility complex-matched transplant model, we evaluated the ability of TK+TC to generate a GVL effect and the ability to control GVHD in experiments where we varied the dose of TK+TC, ganciclovir (GCV) dose, the start of GCV administration (day 4, 7, 10, 13, 15, or 19) posttransplantation, and the GCV administration route (osmotic pump versus intraperitoneal).

Results: At TK+TC doses in excess of the standard lethal dose (SLD) of unmanipulated T-cells, GCV administration completely (2 x SLD) and partially (4 x SLD) controlled GVHD. Additionally, GVHD remained reversible despite delaying administration of GCV for a week after GVHD developed. Importantly, GVHD was controlled with a 1-log but not 2-log reduction in GCV dose, and this "partial suicide" preserved more circulating TK+TC compared with standard-dose GCV. Survival of leukemia-positive mice receiving TK+TC and GCV was significantly increased compared with control cohorts not receiving GCV or transplanted with unmanipulated T cells, thereby demonstrating a GVL effect.

Conclusion: Retrovirally transduced suicidal lymphocytes generate a potent GVL effect while simultaneously enabling control of GVHD, which results in improved leukemia and GVHD-free survival.

Figure 1

TH1 and TH2 cytokine expression in CD3 purified and TK⁺TC CD4 and CD8 T cells. T cells retrovirally transduced to contain the HSV-TK⁺ gene, as described, or CD3 cells purified from splenic lymphocytes were stimulated with 20 ng/mL PMA, 1 μM ionomycin, and BD Golgistop for 6 hours. The percentage of (A) CD4⁺ cells and (B) CD8⁺ cells producing TH1 cytokines; IL-2 and IFNγ, or TH2 cytokines; IL-4 and IL-10 was measured by flow cytometry. A greater percentage TK⁺TC were found to produce cytokines compared with CD3-purified cells. However, results were only statistically significant when comparing CD4⁺ IL-2 levels and CD8⁺ IL-10 levels. Means and standard errors of three experiments are displayed. (C): CD3 purified and TK⁺TC CD4⁺ and CD8⁺ cells (unstimulated with PMA, ionomycin, or Golgistop) were analyzed for expression levels of surface CD44 and CD62L. CD4⁺ and CD8⁺ TK⁺TC populations were found to contain more cells expressing a memory phenotype (CD62Llo/CD44hi) than CD3-purified cells. Results from one of three representative experiments are displayed.

Figure 2

Ability to control GVHD with GCV after infusion of 2 × 10⁶, 4 × 10⁶, or 8 × 10⁶ TK⁺TC. (A): Mice were given transplants as described in Materials and methods of 1 × 10⁷ bone marrow cells (BM, ●) alone (to serve as a GVHD control in this model as the marrow has insufficient lymphocytes to initiate GVHD) or bone marrow cells plus 2 × 10⁶ unmanipulated T cells (NL, ■) or TK⁺TC (TK, ▲) or TK⁺TC followed by intraperitoneal (IP) administration of GCV on days 10 through 16 (TK⁺GCV, Δ). Results from several experiments are combined for the three control groups and from two experiments for the TK⁺GCV group. The number of mice treated is shown in parentheses by each curve. The survival time of mice receiving either bone marrow only or TK⁺TC plus GCV was significantly longer compared with mice receiving either unmanipulated T cells or TK⁺TC without GCV (p < 0.00001 for all four comparisons). (B): Mice were transplanted as described in the Material and methods with 1 × 10⁷ bone marrow cells alone (BM, ●) or along with 4 × 10⁶ unmanipulated T cells (NL, ■), or TK⁺TC (TK, ▲) or TK⁺TC with administration of GCV via an Alzet pump on days 10 through 16 (TK⁺GCV, Δ). Results from several experiments are combined for the three control groups and from two experiments for the group receiving GCV. The number of mice treated is shown in parentheses by each curve. The survival of mice receiving bone marrow only or TK⁺TC plus GCV was significantly longer than that of mice receiving either unmanipulated T cells or TK⁺TC without GCV (p < 0.00002 for all four comparisons). (C): Mice were transplanted as described in Material and methods with 1 × 10⁷ bone marrow cells alone (BM, ●) or along with 8 × 10⁶ or TK⁺TC (TK, ▲) or TK⁺TC with administration of GCV via Alzet pump on days 7 through 13 (TK⁺GCV Days 7–13, □) or on days 10 through 16 (TK⁺GCV Days 10–16, Δ). Results from several experiments are combined for the three control groups and from one experiment for the groups that received GCV. The number of mice treated is shown in parentheses by each curve. The survival time of mice receiving TK⁺TC + GCV on days 10 through 16 was significantly longer than that of mice receiving TK⁺TC without GCV (p < 0.0001) and was statistically similar to that of bone marrow only controls (p = 0.25). The survival time of mice receiving TK⁺TC plus GCV on days 7 through 13 was statistically similar to that of mice receiving TK⁺TC without GCV (p = 0.63) or those receiving TK⁺TC + GCV on days 10 through 16 (p = 0.17) but was inferior to that of the bone marrow–only controls (p = 0.05).

Figure 3

Effect of route of administration and timing on the ability to control GVHD. (A): Mice were given transplants as described in Materials and methods of 1 × 10⁷ bone marrow cells plus 2 × 10⁶ TK⁺TC (TK) or 2 × 10⁶ TK⁺TC followed by the intraperitoneal administration of GCV daily for 7 days starting on days 4 (○), 7 (□), 10 (Δ), or 13 (+). The survival time of all treated groups was significantly longer than that of the untreated group (p = 0.005 [day 4], p = 0.00002 [day 7], p < 0.00001 [day 10], or p = 0.005 [day 13]), even though some mice in the day 13 group had died of GVHD before the initiation of GCV. (B): Mice were given transplants as described in Materials and methods of 1 × 10⁷ bone marrow cells plus 2 × 10⁶ TK⁺TC (▲) or TK⁺TC followed by administration of GCV via an Alzet pump for 7 days beginning on day 7 (Δ) or day 15 (□). The survival time of both treated groups was significantly longer than that of the untreated control (p = 0.005 for day 7 and p = 0.001 for day 15). For the sake of clarity, the survival times of mice that received either bone marrow alone or unmanipulated lymphocytes are not shown here but are shown in Figure 1. Results from several experiments are combined. The number of mice treated is shown in parentheses by each curve.

Figure 4

Control of GVHD by GCV administration using the GVHD score and luciferase imaging. (A): Mice were scored for GVHD as described in Materials and methods. The daily score for one of the experiments described in Figure 2 is shown. A rapid decline in the GVHD score is apparent in the cohort treated with TK⁺TC (TK) that received GCV on days 10 through 16 by intraperitoneal injection. BM, bone marrow; NL, unmanipulated T cells. (B): Mice were transplanted as in the other experiments, but with 4 × 10⁶ HSV-TK⁺ T cells cotransduced with fiber-modified adenoviral vector expressing firefly luciferase. Representative control (top row) and GCV-treated days 7 to 13 (bottom row) mice were imaged as described on days 3, 10, and 18, corresponding to early GVHD pre-GCV, mid-GCV, and post-GCV treatment time points and sacrificed on day 19. (C): Imaging of the internal organs from two representative mice sacrificed on day 19. Organs are labeled as follows: I, intestines; LHL, lung–heart–lung; L, liver; LK, left kidney; RK, right kidney; S, spleen. (D): Relative luminescence of the GCV-treated mice compared with the control mice at days 3, 10, and 18 is shown. The period of GCV administration is shown by the gray rectangle. Signal intensity of the GCV-treated mice relative to the control mice was 1.97, 0.11, and 0.115 at days 3, 10, and 18 days, respectively, for the whole body images and 0.14 (liver), 0.17 (lungs), 0.26 (spleen) 0.30(kidneys) 0.55(intestines) for the organs.

Figure 5

Effect of GCV dose and duration on ability to control GVHD and the number of circulating TK⁺TC. (A): Mice were given transplants as described in Materials and methods of 1 × 10⁷ bone marrow cells alone (●) or 1 × 10⁷ bone marrow cells and 4 × 10⁶ TK⁺TC, with or without GCV (▲) and administered over 7 days via an Alzet pump from days 7 through 13 after transplantation at a total dose of 5 mg (Δ), 0.5 mg (□), or 0.05 mg (◆). The survival times of groups treated with a total dose of 5 mg or 0.5 mg were significantly longer than that of the untreated group (p < 0.00001 and p = 0.0002, respectively) and similar to that of the bone marrow–only control group (p = 0.74 and p = 0.76 respectively). For the cohort receiving a total dose of 0.05 mg, the survival time was similar to that of the untreated control group (p = 0.18) and worse than that of the bone marrow–only control group (p = 0.04). (B): Mice were given transplants as described in Materials and methods of 1 × 10⁷ bone marrow (BM) cells alone (●) or 1 × 10⁷ bone marrow cells with 4 × 10⁶ TK⁺TC (TK) without GCV (▲) or with intraperitoneal (I.P.) administration of 2 mg/d of GCV on days 7 and 8 (Δ), days 10 and 11 (□), or 13 and 14 (◇). The survival times of all three treatment groups were significantly longer than that of the untreated control group (p ≤ 0.005 for all comparisons) and were statistically similar to that of the bone marrow–only controls (p ≥ 0.23 for all comparisons). (C): Approximately 100 μL of peripheral blood was collected on day 21 after transplantation from the tail vein of two mice given transplants as described elsewhere and treated with 5 or 0.5 mg of GCV via an Alzet pump from days 7 through 13 after transplantation. The percentage of circulating cells expressing LNGFR (shaded contour) was assessed by flow cytometry, and the percentage positive cells was determined by comparison with the immunoglobulin G isotype control (unshaded contour).

Figure 6

TK⁺TC generate a GVL effect. All mice were given transplants as described in Materials and methods with 1 × 10⁷ bone marrow (BM) cells alone (●) or with 50 M1 leukemia cells (Leuk) (O) or with 4 × 10⁶ normal lymphocytes and 50 M1 leukemia cells (NL + Leuk, □) or with TK⁺TC (TK) without GCV (Δ), with GCV administered via an Alzet pump on days 10 through 16 (▲) after transplantation, with 50 M1 leukemia cells (Leuk) (◇) or with 50 leukemia cells and GCV administered via an Alzet pump on days 10 through 16 at a dose of 5 mg/wk (◆) or 0.5 mg/wk (*). Mice receiving TK⁺TC, leukemia cells, and GCV had a significantly longer survival times than did mice treated with TK⁺TC and leukemia cells without GCV (p = 0.002) but significantly shorter survival times than did mice receiving bone marrow alone (p = 0.02) or TK⁺TC and GCV without leukemia cells (p = 0.02). Results from seven experiments are combined. The number of mice treated is shown in stated in the legend. In the TK + Leuk + GCV 5-mg group, two mice are censored at 63 and 73 days and seven others survived >100 days.

Figure 7

Long-term expression of TK⁺TC is not altered by a leukemic challenge. Approximately 100 μL of peripheral blood was collected on day 134 after transplantation from the tail vein of two mice given transplants of bone marrow and 4 × 10⁶ TK⁺TC (TK) ± 50 M1 leukemia cells (Leuk) and treated with 5 mg of GCV via an Alzet pump from days 10 through 16 after transplantation. The percentage of circulating cells expressing LNGFR (shaded contour) was assessed by flow cytometry, and the percentage of positive cells was determined by comparison with the immunoglobulin G isotype control (unshaded contour). There were no control mice (TK⁺TC but no GCV) that survived a similar length of time for comparison.