Adoptive Transfer of Natural Killer (NK) Cells to Prevent Gvhd and Enhance GVT Effects After Allogeneic Hematopoietic Cell Transplantation (HCT): The Timing of Donor NK Cell Infusions Critically Impacts Transplant Outcome.

Abstract 786 Murine models have shown infusions of donor NK cells from hematopoietic stem cell transplant donors can prevent GVHD while simultaneously mediating a graft-vs-tumor (GVT) effect. Reduction of GVHD by alloreactive NK cells can be mediated indirectly through the eradication of host antigen presenting cells (APC) or directly by NK cell killing of alloreactive T cells. To assess how the timing of NK cell administration impacts these effects, donor NK cells from B10.d2 (H-2d) mice (1-2×106 cells) were infused into MHC-matched BALB/c (H-2d) recipients following lethal irradiation (950cGy) at one of the following time-points: 1) two days prior to a T cell replete (TR) HCT to target host APC or 2) at the time of a TR-HCT or 3) five days following a TR-HCT to target both host APC and in vivo primed alloreactive donor T cells. We also evaluated whether donor NK cells given on the day of a T cell deplete (TD) HCT could be used to prevent GVHD in mice that subsequently received a delayed donor lymphocyte infusion (DLI) given four days following HCT. Administration of donor NK cells two days prior to allogeneic TR-HCT did not result in a reduction of GVHD (figure). In contrast, the administration of NK cells given either at the time of a TR-HCT or five days following a TR-HCT reduced the incidence of GVHD (GVHD incidence 70% (p=0.2) and 40% (p=0.01) respectively) compared to controls that did not receive NK cells following a TR-HCT (GVHD incidence 100%). Similarly, mice that received a TD-HCT followed by a DLI on day four had a significantly lower incidence of GVHD when NK cells were infused on the day of transplantation compared to controls that did not receive donor NK cells (GVHD incidence 40% vs 100% respectively, p=0.01). Using bioluminescence imaging, we next investigated the impact of the timing of NK cell infusions on GVT effects in tumor bearing mice. Luciferase transduced RENCA tumors were injected intravenously into BALB/c mice ten days prior to allogeneic HCT. Tumor progression was significantly delayed in recipients of a TR-HCT when NK cells were infused on day five compared to when NK cells were infused two days prior to or at the same day of a TR-HCT (tumor doubling times 22.9 days, 7.6 days and 8.5 days respectively; p=0.03) (figure). This delay in tumor progression correlated with a significant improvement in overall survival; recipients of a TR-HCT given NK cells on day five had significantly longer survival compared to recipients of TR-HCT that did not receive NK cells (median survival 54±14 days vs 44±9; p=0.008), whereas infusion of NK cells prior to or concomitant with a TR-HCT did not significantly prolong survival (median survival 49±4 days and 50±12 days respectively; p=0.23). A comparable delay in tumor progression and longer survival was observed in mice that received a TD-HCT followed by a DLI on day four when NK cells were infused at the time of transplant compared to controls not receiving NK cells (tumor doubling time 19.7 days vs 7.2 days and survival 62±16 days vs. 50±9 days respectively; p=0.07). In conclusion, these results show that the timing of adoptive donor NK cell transfer has a critical impact on the ability of NK cells to prevent GVHD and enhance GVT effects following both T-cell replete and T-cell depleted allogeneic HCT. Following a TR-HCT, a delayed add-back of NK cells maximizes GVHD reducing and anti-tumor effects of adoptively transferred donor NK cells. Disclosures: No relevant conflicts of interest to declare.