The Role of Pretransplant Conditioning On Graft-Versus-Leukemia Effects: Migration Matters.

Abstract 3563

Poster Board III-500

The therapeutic potential of allogeneic hematopoietic stem cell transplantation (HSCT) relies on the graft-versus-leukemia (GVL) effect to eradicate residual tumor cells by immunologic mechanisms. However, the relationship of conditioning intensity to GVL effect has not been clearly established independent of immunosuppression or the tolerance induced by mixed donor-host chimerism. Using a murine allogeneic HSCT model, we have compared two total body irradiation (TBI) doses (1,300 vs. 900 cGy), both of which provided complete donor engraftment and elimination of host lympho-hematopoetic cells.

We used C57BL/6 (H-2^b) → B6D2F1 (H-2^b/d) model of GVHD, which differ at major and minor histocompatibility loci, to address the role of conditioning intensity on the GVL effect. Lethally irradiated (either 900 or 1300 cGy) recipient mice were transplanted with either C57BL/6 (allogeneic) or B6D2F1 (syngeneic) bone marrow (5 × 10⁶) and spleen T cells (1 × 10⁶) on day 0 and then P815 (H-2^d) mastocytoma cells (1 × 10⁶) injected subcutaneously on day 1 to generate a GVL model.

As expected, GVHD morbidity after the higher TBI dose was aggravated compared to the lower TBI dose (P<.05). Among the syngeneic recipients, the injection of P815 cells into the recipient skin led to progressive tumor growth and death of about 100% 21 days after transplant regardless of the TBI dose. In contrast, tumor growth was remarkably suppressed and tumor death was not observed in the allogeneic recipients. Surprisingly, tumors in the allogeneic recipients receiving 1300 cGy TBI exhibited markedly delayed growth in vivo compared to those with 900 cGy (tumor volume on day 42, 428 vs. 8735mm³, P<.01), which was associated with an increase in the in vivo cytotoxicity using comparing the clearance of infused allogeneic B cells labeled with CFSE reflecting the enhanced alloimmune reactivity. To ask whether the diminished GVL effect after the lower TBI dose was due to reduced production of inflammatory cytokines, we measured the levels of TNF-α or IFN-γ in recipient sera on days 6, 28 and 42 after transplantation and did not find any significant difference according to the intensity of radiation dose (P>.05). In parallel, the in vitro P815-specific TNF-α or IFN-γ responses of splenocytes were comparable between the two doses. The percentages of donor T cells to undergo proliferation or apoptosis in response to alloantigens in vivo between the two TBI doses also were comparable (P>.05). Collectively, these data indicate that the impaired ability of alloreacive T cells to inhibit tumor growth after the lower TBI dose was not attributed to an intrinsic defect in T-cell expansion and activation. We next analyzed the spleen for the number of donor CD4⁺ and CD8⁺ T cells and observed no difference between the two TBI doses. In contrast to spleen, the number of CD8⁺ but not CD4⁺ T cells from the recipients that had received 1300 cGy was significantly increased in the skin (P<05). The effector function of donor CD8⁺ and CD4⁺ cells in both spleen and tumor tissue was examined by intracellular staining for IFN-γ. In the spleen, the percentages of CD8⁺ and CD4⁺ T cells expressing IFN-γ were not different between the two TBI doses. (5.9% vs 4.8%, P>.05, and 7.6% vs. 6.5%, P>.05 respectively) By contrast, 45.5% and 50.3% of CD8⁺ and CD4⁺ T cells, respectively, isolated from the tumor tissue of recipients receiving the higher TBI dose were IFN-γ; secreting cells, whereas only 25.5% and 16.3% of those cells from the tumor tissue of recipients treated with the lower dose showed this phenotype (P<.01 and <.05, respectively). After the higher TBI dose, secondary lymphoid organ homing receptors including CD62L and CCR7 were down-regulated on donor CD8⁺ T cells while CD44 expression was up-regulated compared to the lower TBI dose, which may facilitate migration to the tumor sites.

In summary, the higher TBI dose (1300 vs. 900 cGy) resulted in significantly enhanced GVL effect, and the alterations in effector T cell trafficking into tumor tissue are the most likely mechanism. Moreover, T-cell activation and function were largely comparable between these conditioning regimens. This provides the rationale for targeting T cell trafficking by inflammation, possibly in combination with integrin or chemokine receptor agonists as a new therapeutic approach in leukemia relapse after allogeneic HSCT.