Table 1.
Selected Clinical Trials Investigating NK Alloreactivity
| Reference | Disease | Number of patients | Conditioning | TCD | Graft source and composition | NK alloreactivity* | GVHD prophylaxis | Engraftment failure | Acute GVHD grade II +/Chronic GVHD | Infection | TRM | RFS | OS | Benefit from NK alloreactivity? |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Haploidentical | ||||||||||||||
| Ruggeri et al., 2007 [16] | AML | 112 | MA | Ex vivo | PB; 15 × 106/kg CD34, 3 × 104/kg CD3 | 1, 2, 3 | 0 | 6% vs 10% | 10%, NS/NR | 38% fatal infections | 43% | 67% vs 18% | NR | Yes |
| Leung et al., 2004 [86] | AML, ALL; pediatric | 36 | NR | CD34+ selection | PB; < 3 × 104/kg CD3 | 1, 2, 4 | 0 | NR | NR/NR | NR | NR | 13% vs 54% relapse rate | NR | Yes, for relapse, “missing ligand” model |
| Lang 2004 [134] | Various | 63 | MA | CD34 or CD133 selected + ATG | PB; 19.5 × 106/kg MNC, < 2.5 × 104/kg CD3 | 1 | 0 | 17% | 7%/13% | 17% fatal infections | 27% | NR | 42% | No (equivalent to historical matched unrelated donor) |
| HLA-identical related | ||||||||||||||
| Hsu et al., 2005 [47] | AML, CML, ALL, MDS | 178 | MA | Ex vivo | BM; 9× 105/kg CD3 | 2 | Yes | 0% | NS/NS | NR | NR | × 0.41 relapse (AML, MDS) | × 0.52 risk | Yes |
| Cook 2004 [135] | Various | 220 | MA/RIC | NR | NR | 1, 2 | Yes | NR | NS/NR | NR | NR | NS | 31.6%) vs 56.1% (4 years) | No; worse survival for myeloid patients with C2/C2 and KIR2DS2 donor |
| Unrelated | ||||||||||||||
| Giebel 2003 [136] | Various | 130 | MA | ATG | BM; 4.3 × 108/kg MNC | 1 | Yes | 0% vs 4% | 0% vs 15% (grade III-IV/NS | NR | 6% vs 40% | Relapse 6% vs 21% | 87% vs 48% (4.5 years | Yes |
| Kroger et al., 2006 [104] | AML, CML, ALL, MDS | 142 | MA | ATG | PB/BM | 1, 2 | Yes | 0% | NS/NS | Increased | × 2.2 risk if alloreactive | × 3 relapse risk (activating KIR) | × 0.5 unless donors are KIR haplotype A | No; ligand/ligand model |
| Cooley et al., 2008 [59] | AML | 448 | MA | No | PB/BM | 1, 2 | NR | NR | NS/× 1.5 risk if activating haplotype | NR | NS | × 2 RFS (activating KIR) | × 1.5 with higher number of activating KIR (3 years) | Y for donors with group B KIR haplotype |
| Davies 2002 [137] | Various | 175 | MA | Ex vivo, minority | BM; 2 × 108/kg MNC | 1 | Yes | NS | NS/NR | NR | NR | 9%-12% at 5 years (NS) | NS (whole group); × 0.5 (myeloid) | No |
| Farag 2006 [138] | AML, CML MDS | 1571 | MA | Minority ex vivo TCD | BM | 1 | NR | NS | NS/NS | NR | × 1.95 risk | NS | × 0.5 risk | No |
| Yabe et al., 2008 [132] | Various | 1489 | MA | ATG, minority | BM | 1, 2 | Yes | NR | × 1.7 risk; increased with 2DS2 gene in donor/NR | NR | NR | NS | × 1.93 risk | No (worse GVHD) |
| Miller 2007 [139] | AML, CML MDS | 2062 | NR | NR | NR | 1 | NR | NR | 44% vs 30% (late-phase CML)/NR | NR | NR | × 0.54 relapse in early disease | NR | Yes, for early myeloid disease |
| Bornhauser et al, 2004 [99] | AML, CML, MDS | 118 | MA | ATG | PB/BM; 4 × l06/kg CD34 | 1 | NR | 10%, NS | 46% vs 69%, NS/NS | NR | NR | Relapse 60% vs 35% | NS | No |
RFS indicates relapse-free survival; OS, overall survival; MA, myeloablative conditioning; RIC, reduced-intensity conditioning; ATG, antithymocyte globulin; MNC, mononuclear cell; haploA, KIR haplotype A; C2, HLA-C2 alleles; NS, no significant difference; NR, not reported.
*
Model used to define NK alloreactivity:
1, ligand/ligand (HLA typing of donor and recipient).
2, receptor/ligand (KIR genotyping or phenotyping of donor).
3, specific cytotoxicity assay of donor NK cells against recipient cells.
4, nonspecific cytotoxicity assay of donor NK cells against NK-susceptible targets (eg, K562 cell line).