The number of CD56dim NK cells in the graft has a major impact on risk of disease relapse following allo-HSCT.

The graft-versus-leukemia (GVL) effect of allogeneic hemopoietic stem cell transplantation (allo-HSCT) is mediated by the donor immune system and acts to decrease the rate of disease relapse. Although studies of posttransplant immune reconstitution have identified correlates of clinical outcome, the number and profile of mature immune cells infused with the stem cell graft is also likely to be an important determinant and has been relatively poorly studied. We characterized immune cells within the stem cell graft of 107 patients who underwent T-cell-depleted allo-HSCT and related this to clinical outcome. The number of natural killer (NK) cells and T cells that were infused varied markedly between patients, but T-cell dose was not an important factor in subsequent outcome. In contrast, the number of NK cells was a powerful determinant of the risk of disease relapse. Patients who received an NK cell dose below the median level of 6.3 × 106 cells per kg had a relapse rate of 40% at 2 years posttransplant compared with only 6% for those whose stem cell graft contained a dose above this value. Analysis of NK subsets showed that this effect was mediated primarily by the CD56dim population of mature effector cells and that high-level expression of the activatory protein DNAM on donor NK cells was also strongly protective. These observations offer important insights into the mechanism of GVL and suggest that optimization studies of the number of NK cells within the stem cell graft should be considered as a means to reduce disease relapse.