Large Scale Ex Vivo GMP Expanded, Activated Human T Cells Consistently Induce Lethal GvHD in a Mouse Xenotransplant Model - A New Way To Study Treatments for Acute GvHD.

Donor lymphocyte infusions have become increasingly valuable for the treatment of hematologic diseases. In the past it has been reported that ex vivo activated and expanded human (hu) T lymphocytes, compared to fresh, non activated donor lymphocytes, have a markedly reduced ability to elicit normal immune functions in vivo. To overcome this problem, an improved method of ex vivo T cell activation and expansion and a model for in vivo testing of the function of cultured T cells is needed. We developed a more physiological, closed system GMP T cell stimulation protocol using clinical grade magnetic beads coated with CD3 and CD28 antibodies (Xcyte Therapies Inc., Seattle, WA), IL-2 at a low concentration added 24 hours after the initial culture setup and the new GMP grade serumfree Stemline T cell medium (Sigma, St. Louis, MO). This medium allowed the elimination of donor variability, which we experienced in T cell expansions with other serumfree media. Using our NOD SCID/β2m null mouse model of T cell expansion, we demonstrated that high concentrations of IL-2 (500 U/ml, as previously used in clinical T cell expansions) impair the in vivo functionality of expanded T cells. We therefore lowered the IL-2 concentration to 50 U/ml during culture. A 3 fold cell expansion after 4 days, and a 10 fold cell expansion after 8 days of culture could be observed. The CD4 and CD8 ratios were 1.6 ± 0.5 at the start, 1.7 ± 0.4 four days post activation and 1.1 ± 0.2 eight days post-activation, with viability greater than 95%. We conditioned NOD SCID/β2m null mice with 300 cGy of TBI and injected them retroorbitally with 1 x 10⁷ non activated (n=8), 4 day activated (n=9), or 8 day activated hu T cells (n=5). Mice that received 4 or 8 day activated huT cells exhibited weight loss, high hu T cell engraftment, and high hu IFNγ serum levels. All these animals exhibited lethal GvHD with median survivals of 15 and 12 days, respectively. Activated cells outperformed non activated hu T cells in their GvHD potential since not all mice injected with non activated cells developed lethal GvHD. Interestingly, mice which received 4 day activated cells exhibited significantly increased percentages of hu T cells in the blood, spleen, liver and lung, as compared to animals that received non activated hu T cells. With non activated hu T cells, preferential expansion of hu CD4+ cells was seen in mouse organs; with activated hu T cells the CD4/CD8 input ratio was preserved. Although 8 day activated hu T cells also elicited lethal GvHD in all animals, we did not observe the same pattern of engraftment and organ infiltration as with 4 day activated cells. 4 day activated hu T cells produced significantly higher engraftment, hu IFNγ production, and destruction of mouse tissue compared to 8 day activated hu T cells. For the first time, these results demonstrate consistent lethal GvHD elicited in a mouse xenotransplant model by 10⁷ex vivo activated, expanded hu T cells produced in a closed system, serumfree GMP manufacturing process outperforming non activated hu T cells.