Genetic Modification of Multi Leukemia Antigen-Specific Cytotoxic T Lymphocytes (CTL) to Enhance In Vivo Safety and Persistency

Abstract 644

Adoptive transfer of ex-vivo expanded leukemia-specific T cells has been clinically evaluated for its ability to produce anti tumor effects without causing graft-versus-host disease (GVHD). Previous protocols, however, allowed the expansion only of HLA class I-restricted CD8 T cells, which were short-lived and permitted tumor immune escape since they recognized only a single epitope/antigen target.

To minimize tumor immune escape and extend this therapeutic alternative to more patients irrespective of HLA type we have developed a novel strategy to produce single cultures of in vitro generated T cell lines containing tumor-cytotoxic T cells from both CD4+ and CD8+ populations with specificity for a multiplicity of epitopes on several tumor associated antigens (TAAs) frequently expressed by myeloid leukemias.

Tumor-directed T cells were activated in vitro using dendritic cells (DCs) loaded with peptide libraries (pepmixes) spanning Proteinase 3 (Pr3), PRAME, and WT1 and expanded in the presence of a cocktail of Th1-polarizing, pro-survival and pro-proliferative cytokines (IL7, 12, 15 and 27). This approach consistently generated TAA-specific cytotoxic T lymphocytes (CTLs) with simultaneous reactivity against PRAME (mean 651+/−69 spot forming units (SFU)/1×10⁶ CTLs), WT1 (mean 156+/−20 SFU/1×10⁶ CTLs) and PR3 (mean 52.3+/−8.8 SFU/1×10⁶ CTLs) in IFNg ELIspot assays (n=10). These lines were polyclonal, comprising antigen-specific CD4+ (mean 32.5+/−2.3%) and CD8+ (mean 38.3+/−4%) T cells. Functionally, these CTLs killed autologous PHA blasts pulsed with PRAME (mean specific lysis 67+/−5% SEM), WT-1 (mean 54+/− 5%) and PR3 pepmixes (mean 16+/−6%, n=6) (E:T of 40:1). In addition, TAA CTLs killed the partially (1-3 alleles) HLA class I or HLA class II matched whole antigen-expressing AML cell line THP1 in coculture assays (n=5).

The in vivo anti-tumor activity of multiTAA-CTLs was tested using a SCID mouse tumor model. After subcutaneous engraftment of FFLuc transduced THP-1 cells, mice were treated with either 1×10⁷ multiTAA-CTLs or control EBV-CTLs generated from the same partially HLA-matched donor (sharing HLA-A2, DRB1-01,DQB1-05) plus 1000U of IL2 (5 times per week). Tumor signals as measured by bioluminescence using an in vivo imaging system rapidly increased in untreated and EBV-CTL treated groups, whereas mice treated with TAA specific CTLs were able to control tumor growth (p<0.05) and significantly improve overall survival (log rank p=0.008).

Next, we transduced multiTAA-CTL with a retroviral vector (SFG-hIL-7R/iCasp9) encoding IL7Ralpha and the inducible suicide gene iCaspase9. This genetic modification improved the in vivo persistence, anti-tumor potency of multiTAA-CTL by restoring their response to homeostatic IL7. Furthermore we improve safety with the potential to control unwanted proliferation of the transgenic cells and to treat potentially occurring autoimmune disease caused by targeting self antigens through the administration of a small molecule chemical inducer of dimerization (CID) (AP20187) which activates the suicide gene.

Transduction of multiTAA-CTLs did not adversely affect the antigen specificity of the product as evaluated by IFNg ELIspot (528.5 SFU/1×10⁶ CTL transduced vs. 573.9 SFU/1×10⁶ CTL non transduced, PRAME, 86.7 vs.46.5 SFU/1×10⁶ CTL, Pr3 and 129.2 vs 130.2 SFU/1×10⁶ CTL and WT1 transduced vs. non transduced); and cytolytic activity against peptide-pulsed PHA blasts as well as whole antigen presenting tumor cell lines was maintained. In the presence of IL-7, transgenic multiTAA-CTLs expanded significantly (transduced cells increased from 55%+/1% to 89%+/−1%). Such expansion was not observed when these cells were cultured in the presence of IL-15 (decrease 55%+/1% to 34.6%+/−3.5%) (n=8). Proliferation assays confirmed that only IL7/casp9 engineered multiTAA-CTLs were able to proliferate in the presence of IL-7 (84638 CPM vs. 14923 CPM transduced vs non-transduced). Upon activation of iCaspase9 by of CID (AP20187), the proliferation of multiTAA-CTLs was halted and a decrease of transgenic cells from 89% to 3% with was detected within 24 hrs (n=7).

Hence CTLs can be prepared that simultaneously target multiple TAA antigens expressed by myeloid malignancies, reducing the risk of tumor immune escape. In addition these tumor CTLs can be genetically modified to safely improve their in vivo survival, and persistence.