Transplant strategies involving natural killer (NK) cell alloreactivity (KIR-ligand mismatch [KIR-L mismatch]) have demonstrated superior outcomes for patients receiving T cell depleted HLA haploidentical hematopoietic stem cell allografts. It is unknown whether KIR-L mismatch has a similar effect in recipients of partially HLA-matched umbilical cord blood (UCB) grafts which contain comparatively few T-cells. We examined the clinical impact of KIR-L mismatch in 257 UCB recipients treated with either a myeloablative (n=155) or reduced intensity (n=102) regimen. After myeloablative conditioning, KIR-L mismatch had no demonstrable effect on grades III–IV acute GVHD (17% [CI, 6–28%] vs. 17% [CI, 10–24], p=.97), transplant-related mortality (TRM) (27% [CI, 14–40%] vs. 18% [CI, 11–25%], p=.19), relapse at 2 yrs (18% [95%CI, 6–30%] vs. 28% [95%CI, 19–27%], p=.37) and survival at 3 yrs (50% [CI, 32–68%] vs. 57% [CI, 47–67%], p=.46). In contrast, following reduced intensity conditioning when the engrafting unit was KIR-L mismatched there was a significantly higher incidence of grades III–IV acute GVHD (42% [CI, 27–59) vs. 13% [CI, 5–21], p < .01). Multivariate analysis confirmed NK cell alloreactivity as the only predictive factor associated with severe acute GVHD was (RR 1.8, CI [1.1–2.9]; p=.02). TRM was higher (27% [CI, 12–42%] vs. 12% [CI, 5–19%], p=.03) and three year survival was poorer (32% [CI, 15–59%] vs. 52% [CI, 47–67%], p=.03) when the engrafting unit was KIR-L mismatched, but relapse was unaffected (39% [95%CI, 21–57%] vs. 47% [95%CI, 34–60%], p=.72). KIR-L mismatch in recipients of a RIC UCB transplant was associated with a significant increased risk of death (RR 1.8, 95%CI, 1.0–3.1, p=.05). This data identify divergent effects of NK cell alloreactivity which are unmasked when comparing myeloablative versus RIC transplant platforms. We conclude that KIR-L mismatch should be avoided in recipients of a reduced intensity UCB transplant.
Disclosures: Noreen:National Marrow Donor Program: Consultancy.