No abstract

Cytomegalovirus (CMV) reactivation was analyzed in 92 recipients of allogeneic hematopoietic stem cell transplantation (HSCT) in relation to the proportion of CD4(+) lymphocytes in blood and a microsatellite polymorphism within the first intron of the interferon-gamma (IFNG) gene. CMV reactivation was found in 50% of the HSCT recipients; in 30% of these individuals, the level of CMV copies exceeded 100 per 10(5) peripheral blood (PB) cells on at least one occasion during the 100-day post-HSCT observation period. This high CMV copy level was most frequently found between 31 and 60 days post-HSCT (P = .021). Patients with > or = 100 CMV copies/10(5) cells were characterized by poorer overall survival (OS) compared with those lacking CMV copies or having < 100 CMV copies/10(5) cells (P = .04), and they suffered from severe post-HSCT complications, including acute graft-versus-host disease (aGVHD) and relapse. Thus, patients with > or = 100 CMV copies/10(5) cells were designated as having clinically significant CMV reactivation. Patients with < 10% CD4(+) lymphocytes had a higher number of CMV DNA copies than those with higher proportions of CD4(+) lymphocytes (0.62 vs 0.21, P = .001; mean +/- SEM, 4422 +/- 1667 vs 937 +/- 662 CMV copies/10(5) cells, P < .001, for the proportion of cases with reactivation and numbers of copies, respectively). Similarly, patients carrying 2 IFNG 13-CA-repeat alleles (homozygotes) had more frequent CMV reactivation (0.50 vs 0.26; P = .039) and a higher CMV load (4111 +/- 1699 vs 950+/-591 CMV copies/10(5) cells; P = .041) compared with those with other IFNG microsatellite allele constellations. Multivariate analysis demonstrated that the IFNG 13-CA-repeat homozygous genotype (odds ratio [OR] = 0.221; P = .044), a low proportion of CD4(+) lymphocytes (OR = 0.276; P = .050), and a lack of optimal (10/10 alleles) donor-recipient HLA match (OR = 15.19; P = .006) were independent risk factors for CMV reactivation with a high number of copies.

Some cancers treated with allogeneic hematopoietic stem cell transplantation (HSCT) are sensitive to natural killer cell (NK) reactivity. NK function depends on activating and inhibitory receptors and is modified by NK education/licensing effect and mediated by coexpression of inhibitory killer-cell immunoglobulin-like receptor (KIR) and its corresponding HLA I ligand. We assessed activating KIR (aKIR)-based HLA I-dependent education capacity in donor NKs in 285 patients with hematological malignancies after HSCT from unrelated donors. We found significantly adverse progression-free survival (PFS) and time to progression (TTP) in patients who received transplant from donors with NKs educated by C1:KIR2DS2/3, C2:KIR2DS1, or Bw4:KIR3DS1 pairs (for PFS: hazard ratio [HR], 1.70; P = .0020, Pcorr = .0039; HR, 1.54; P = .020, Pcorr = .039; HR, 1.51; P = .020, Pcorr = .040; and for TTP: HR, 1.82; P = .049, Pcorr = .096; HR, 1.72; P = .096, Pcorr = .18; and HR, 1.65; P = .11, Pcorr = .20, respectively). Reduced PFS and TTP were significantly dependent on the number of aKIR-based education systems in donors (HR, 1.36; P = .00031, Pcorr = .00062; and HR, 1.43; P = .019, Pcorr = .038). Furthermore, the PFS and TTP were strongly adverse in patients with missing HLA ligand cognate with educating aKIR-HLA pair in donor (HR, 3.25; P = .00022, Pcorr = .00045; and HR, 3.82; P = .027, Pcorr = .054). Together, these data suggest important qualitative and quantitative role of donor NK education via aKIR-cognate HLA ligand pairs in the outcome of HSCT. Avoiding the selection of transplant donors with high numbers of aKIR-HLA-based education systems, especially for recipients with missing cognate ligand, is advisable.