Adoptive Transfer of Primed T Cells Mediates a Graft-Versus-Leukemia Effect against Pediatric ALL.

Pediatric Acute Lymphoblastic Leukemia (ALL) is the most common childhood malignancy. While current upfront therapy will cure over 80% of patients, treating relapse remains a major challenge. Blood or marrow transplantation (BMT) offers a therapeutic option but approximately 50% of patients undergoing BMT will not survive with relapse being the most common cause of death. Donor lymphocyte infusions (DLIs) have demonstrated efficacy in myeloid leukemia but are much less effective for ALL. Thus, we explored strategies to improve on the potency of the graft versus leukemia response against pediatric ALL using a murine model developed from E2a-PBX1 transgenic mice that express a recurring translocation t(1;19) present in approximately 5% of all pediatric ALL.

E2a-PBX1 was administered intravenously (IV). Irradiated E2a-PBX1 was administered intraperitoneally for immunization both in the vaccine protection model and to prime DLI donors in the BMT models. Transplanted mice were lethally irradiated and given T-cell-depleted bone marrow cells followed by purified T cells as a DLI. In the syngeneic model, C57Bl/6 mice were used as both recipients and donors so that both donor cells were syngeneic to the tumor. In the allogeneic model, C57Bl/6 recipients received bone marrow and T cells from MHC-matched C3H.SW donors that are minor antigen-mismatched to both tumor and recipient, analogous to the clinical BMT setting.

We confirmed that E2a-PBX1 is a pre-B cell ALL that expresses B220, BP-1, CD43 and the alpha chain of the IL-7 receptor. The distribution of leukemia is similar to that observed in humans with bone marrow, lymph node, liver, spleen and central nervous system involvement confirmed by necropsy and imaging following injection of luciferase-expressing cells. Amazingly, E2a-PBX1 is lethal with as few as with as 1×10² cells (5/5 mice) in sublethally irradiated mice (250 cGy) and 1×10⁴ cells in unirradiated mice (4/5 mice). Vaccination prior to tumor challenge protected 90-100% of mice from leukemia development. Mice that successfully rejected E2a-PBX1 were protected against rechallenge with a higher cell dose. Antibody depletion of either CD4+ or CD8+ T cells did not affect vaccine-mediated protection. However, depletion of both CD4+ and CD8+ T cells significantly diminished vaccine efficacy (n=7/group, p=0.0063). Interestingly, although data from killer inhibitory receptor (KIR) mismatched BMT would suggest that NK cells are less potent in ALL than in AML, antibody depletion of NK cells with either anti-asialo GM1 or anti-NK1.1 also had a significant impact on survival in our model. We next tested the therapeutic effect of purified T cells injected IV after E2aPBX1 injection. Administration of naive T cells (6×10⁶) following syngeneic BMT was not effective at treating E2a-PBX1. The presence of minor antigen differences between donor T cells and tumor following allogeneic BMT did not increase the efficacy of naïve T cells consistent with the poor response rate to DLI in patients with ALL. However, primed T cell DLI from a donor immunized by irradiated E2a-PBX1 resulted in a statistically significant prolongation of survival in both syngeneic (p=.0067) and allogeneic (p=.008) BMT models. Conclusion: When administered in our allogeneic transplant model, E2a-PBX1 can be used to model relapse of pediatric ALL post-BMT. Naïve T cells (analogous to DLI given post-relapse in the clinic) were ineffective at treating ALL our model. However, priming of the DLI significantly prolonged survival demonstrating that the poor efficacy of DLI in ALL is not due to an inherent resistance of ALL to a T cell response. Our data supports the use of T cells to treat pediatric ALL relapse post-BMT but suggests manipulation of the T cell response will be required.

No relevant conflicts of interest to declare.