Donor KIR3DL1/3DS1 Gene and Recipient Bw4 KIR Ligand as Prognostic Markers for Outcome in Unrelated Hematopoietic Stem Cell Transplantation

Article Outline

• Abstract
• Introduction
• Materials and Methods
  • Patient, Disease, and Transplant Characteristics
  • Genomic DNA Extraction
  • HLA Class I and Class II Genotyping
  • KIR Genotyping
  • Statistical Analysis
• Results
  • Impact of Donor/Recipient KIR Ligand Disparities According to the “Ligand-Ligand” Model on Unrelated HSCT Outcome
  • Impact of “Missing KIR Ligands” in Patients on Unrelated HSCT Outcome
  • Impact of D/R KIR Gene Disparities According to the “Receptor-Receptor” Model on Unrelated HSCT Outcome
  • Impact of Donor KIR Genes and Recipient HLA Class I Ligand Disparities According to the “Receptor-Ligand” Model on Unrelated HSCT Outcome
• Discussion
• Acknowledgments
• References
• Copyright

Given their antileukemic activity, natural killer (NK) cells can alter the outcome of hematopoietic stem cell transplantation (HSCT). The physiologic functions of NK cells are regulated by the interaction of killer immunoglobulin-like receptors (KIR) with specific HLA class I ligands. In the literature, different models based on HLA class I and/or KIR donor (D)/recipient (R) gene disparities are considered as predictors of NK cell alloreactivity. In this retrospective and multicentric French study, we analyzed the clinical impact of the different NK-alloreactivity models in 264 patients who underwent T repleted unrelated HSCT. First, we did not observe that the “KIR ligand-ligand” model had a significant clinical impact on unrelated HSCT outcome, whereas the “missing KIR ligand” model had a significant but limited effect on unrelated HSCT, because only the absence of C1 ligand in patients with myelogenous diseases was associated with a decreased overall survival (OS) (hazard ratio=2.17, P=.005). The “KIR receptor-receptor” and the “KIR receptor-ligand” models seemed the most capable of predicting NK alloreactivity because they had a significant impact on acute graft-versus-host disease (aGVHD) occurrence, OS, and relapse incidence in D/R unrelated pairs. In particular, KIR3DL1 gene mismatches in the GVH direction (D⁺R⁻) and the D KIR3DL1⁺/3DS1⁺ and R Bw4⁻ combination were respectively correlated with the lowest OS in HLA identical pairs (HR=1.99, P =.02) and the highest incidence of relapse in HLA nonidentical D/R unrelated pairs (HR=4.72, P =.03). Overall, our results suggest a detrimental effect of KIR3DL1⁺/3DS1⁺ donor NK cells transplanted into HLA-Bw4⁻ patients in the absence of an educational process via KIR3DL1/HLA-Bw4 interactions.

Key Words: Killer immunoglobulin-like receptors, KIR ligands, KIR genotypes, KIR3DL1, KIR3DS1, HLA-Bw4, Unrelated hematopoietic stem cell transplantation, Graft-versus-host-disease, Survival, Relapse

Back to Article Outline

Introduction 

Hematopoietic stem cell transplantation (HSCT) is currently used for the treatment of a variety of hematologic malignancies. Whereas the best outcome following HSCT is observed between HLA-identical siblings, the alternative of using transplants from unrelated donors is increasing. When a perfectly HLA matched donor is unavailable, selection is based on the best HLA matching considering both class I and class II loci [1].

The efficacy of allogeneic HSCT remains limited by several significant complications, such as failure of the hematopoietic stem cells (HSCs) to engraft, the occurrence of acute graft-versus-host disease (aGVHD), relapse, and the susceptibility of patients to opportunistic infections during the posttransplant immunodeficiency period. However, a beneficial graft-versus-leukemia (GVL) response, developed by donor T lymphocytes or natural killer (NK) cells, could be observed after HSCT [2]. In extensive T cell-depleted HSCT, such as in the case of haploidentical grafts, alloreactive NK cells were able to exert this GVL effect without causing aGVHD [3].

The physiologic functions of NK cells, such as cell cytotoxicity, are governed by a balance between inhibitory and activating receptors including the killer cell immunoglobulin-like receptors (KIR), which are specific for allotypic determinants shared by different HLA-class I molecules (referred to as KIR ligands) [4]. Four different inhibitory KIR seem to play a major role in NK cell alloreactivity because of a more specific recognition of different HLA class I-encoded ligands. HLA-Cw allotypes with asparagine at position 80 (C1 ligands) are recognized by KIR2DL2/2DL3, HLA-Cw allotypes with lysine at position 80 (C2 ligands) are recognized by KIR2DL1, HLA-A, and -B allotypes with a polymorphic sequence motif at position 80-83 (Bw4 motif) are targeted by KIR3DL1, and HLA-A3 and -A11 are recognized by KIR3DL2 5, 6. However, although the ligands and functions of inhibitory KIR receptors are well documented, this is not the case for activating KIR receptors and their ligands, except for KIR2DS1 7, 8, 9. In particular, the activating receptor KIR3DS1, which is encoded as an allele of KIR3DL1, shares more than 97% sequence homology in its extracellular domain with the KIR3DL1 receptor. However, a functional interaction with HLA-A or -B allotypes sharing the Bw4 public epitope has not been demonstrated in vitro 10, 11.

KIR genes are located on chromosome 19q13.4. Until now, 14 functional KIR genes have been characterized [12]. Within the human population, genomic diversity of the KIR region is achieved on several levels. First, KIR gene content varies between individuals who can exhibit from 7 to 14 activating and inhibitory KIR genes [13]. Based on population studies, 2 major KIR haplotype groups, the A and B haplotypes, emerged. The A haplotypes have been defined as containing 7 KIR genes, comprising KIR2DS4 as the only activating KIR gene and the inhibitory KIR3DL3, -2DL3, -2DL1, -2DL4, -3DL1, and -3DL2 genes. In contrast, B haplotypes are more variable and are characterized by the presence of more than 1 activating KIR genes and the absence of the KIR2DS4 gene [14]. Finally, KIR gene polymorphism is the largest contributor to the diversity of the KIR region, with multiple alleles already defined [15].

Because KIR and HLA genes are located on different chromosomes (chromosomes 19 and 6, respectively), matching for HLA genes does not generally result in matched KIR genes in HSCT. Moreover, in some instances HLA class I mismatching may represent a KIR-ligand mismatching, leading potentially to NK alloreactivity. This NK cell alloreactivity appears particularly beneficial in haploidentical [16] and umbilical cord blood transplantations [17], but its influence on the outcome of unmanipulated related and unrelated HSCT for hematologic malignancies remains controversial 18, 19.

In the literature, different models are considered for the prediction of NK alloreactivity, referred to as the “KIR ligand-ligand” model [3], the “missing KIR ligand” model [20], the “KIR receptor-receptor” model [21], and the “KIR receptor-ligand” model [22]. A majority of studies published so far have focused only on KIR-ligand donor (D)/recipient (R) mismatches (“KIR ligand-ligand” model) considering the major HLA class I specificity groups (Bw4, C1, and C2) recognized by some inhibitory KIR. These become potentially beneficial mismatches or “perfect mismatches” only if a donor NK genotype displays the corresponding inhibitory KIRs (ie, KIR2DL1, KIR2DL2/2DL3, and KIR3DL1), and if the corresponding HLA class I (KIR-ligand mismatch) is absent in the recipient. The “KIR receptor-receptor” model takes into account the KIR genotype of both D and R, and, in particular, unravels the clinical impact of inhibitory and activating KIR gene mismatches. The “KIR receptor-ligand” model may represent a more predictive model of potential NK alloreactivity because both the D KIR genotype and the presence or absence of the corresponding R KIR ligand (“missing KIR ligand”) are taken into account. Overall, only a combinatory HLA class I and KIR genetic study should enable a better evaluation of the real impact of NK alloreactivity in HSCT, because the NK-KIR repertoire is not only defined by the combination of inhibitory and activating KIR genes (KIR genotypes), but also by the HLA genotype 23, 24, 25, 26.

In this retrospective and multicentric French study, we analyzed the clinical impact of the 4 defined NK-alloreactivity models in 264 patients who underwent T repleted unrelated HSCT. Our goal was to determine which model is the most accurate in predicting aGVHD occurrence, overall survival (OS), and relapse incidence in HLA identical and nonidentical D/R pairs. Because we previously showed in a preliminary report a detrimental effect of KIR3DL1/3DS1 gene disparities on patient survival in unrelated HSCT [27], we also focused on the clinical relevance of the D KIR3DL1/3DS1 gene together with the absence of HLA-Bw4 ligand in recipients assumingly the follow-up in terms of OS, aGVHD, and relapse.

Back to Article Outline

Materials and Methods 

Patient, Disease, and Transplant Characteristics 

The study design was approved by the institutional review boards of the Société Française de Greffe de Moelle et de Thérapie Cellulaire (SFGM-TC). D/R unrelated pairs were selected from the 13th International Histocompatibility Working Group in Hematopoïetic Stem Cell Transplantation (Pr E. Petersdorf, Seattle, WA), and included complete clinical and biologic data. Patients who received a unrelated transplant for the treatment of a hematologic disorder and who satisfied all of the following criteria were eligible for study: (1) written informed consent obtained from all patients and D by the participating laboratory/transplant center permitting inclusion of HLA, KIR, and clinical data for analysis and reporting; (2) HLA-A, -B, -Cw, -DRB1, and -DQB1 allele typing available for both the patient and the D; (3) KIR genotyping available for both the patient and the D; (4) survival, aGVHD, and relapse incidence data available on the patient.

Two hundred sixty-four patients with a median age of 24.5 years (range: 1-56 years) and who received an unmanipulated bone marrow (BM) graft from an unrelated D were included in this analysis. The grafts were performed in 9 French transplant centers between 1993 and 2003 to treat malignant (n=229) or nonmalignant diseases (n=35). Malignant diseases included acute leukemia (AL) (n=138), chronic myelogenous leukemia (CML) (n=54), myelodysplastic syndrome (MDS) (n=23), myeloproliferative syndrome (n=1), non-Hodgkin lymphoma (NHL) (n=11), Hogdkin disease (HD) (n=1), and myeloma (n=1). All patients were prepared for transplantation with the use of myeloablative (MA) conditioning regimens. The source of grafted cells for all patients was BM without T cell depletion.

The minimal clinical information collected for each patient was aGVHD (date of onset and grade), date of the last visit (alive/deceased status), and relapse. The severity of aGVHD was graded according to international criteria 28, 29. All patients were considered eligible for aGVHD evaluation at day +1 after transplantation. The overall incidence of grade II-IV aGVHD and severe aGVHD (grade III-IV) at 100 days posttransplant was 53.4% (n=141) and 25.8% (n=68), respectively. At 2 years posttransplant, 130 patients had survived (49.2%) and 134 had died. Finally, 55 of 229 (24%) patients suffering from malignant diseases had relapsed at 30 months posttransplant.

The main patient, disease, and transplant characteristics are described in Table 1.

Table 1. Patient, Disease, and Transplant Characteristics of the 264 Unrelated Pairs Included in the Study

Variable	Categories	n (%)
Patient
Diagnosis
	Malignant diseases	229 (86.7)
	Myelogenous diseases	147 (64.0)
	No myelogenous diseases	79 (35.0)
	Not classified	3 (1.0)
	Acute leukemia	138 (52.3)
	Chronic myelogenous leukemia	54 (20.5)
	Myelodysplastic syndrome	23 (8.7)
	Myeloproliferative syndrom	1 (0.4)
	Non-Hodgkin lymphoma	11 (4.2)
	Hodgkin disease	1 (0.4)
	Myeloma	1 (0.4)
	Nonmalignant diseases	35 (13.3)
	Hemoglobinopathy	1 (0.4)
	Aplasia	26 (9.8)
	Inborn errors	3 (1.1)
	Other	5 (1.9)
Transplant:
Year of transplantation	1993-2000	177 (67.0)
	2001-2003	87 (33.0)
Sex of D/R	Male/male	90 (34.0)
	Female/female	61 (23.0)
	Female/male	64 (24.0)
	Male/female	49 (19.0)
CMV status of D/R	Pos/pos	36 (14.0)
	Neg/neg	106 (40.0)
	Pos/neg	51 (19.0)
	Neg/pos	47 (18.0)
	Unknown	24 (9.0)
Preparative regimen	TBI/Cy	186 (70.5)
	Bu/Cy	37 (14.0)
Source of stem cells	unmanipulated bone marrow	264 (100.0)
Acute GVHD incidence	Grades II-IV 100 days posttransplant	141 (53.4)
Survival status	Alive patients at 2 years posttransplant	130 (49.2)
Relapse incidence	At 30 months for malignant diseases	55 (24.0)

D indicates donor; R, recipient; CMV, cytomegalovirus; TBI, total body irradiation; Cy, cyclophosphamide; Bu, busulfan; GVHD, graft-versus-host disease.

Genomic DNA Extraction 

Anticoagulant peripheral blood (PB) samples were obtained from all patients before initiating the conditioning regimen and on the day of transplantation from all Ds. Peripheral blood mononuclear cells (PBMC) from patients and Ds were used as a source of genomic DNA extracted using a classical salting-out method.

HLA Class I and Class II Genotyping 

High-resolution typing for HLA-A, -B, -Cw, -DRB1, and -DQB1 was performed using PCR-SSP amplification kits (OneLambda, Inc., Canoga Park, CA, or Genovision, West Chester, PA, or Dynal, Invitrogen Cergy-Pontoise, France) or by sequence-based typing as previously described [30]. In all cases, an unambiguous 4-digit typing was obtained.

KIR Genotyping 

All patients and Ds were typed for the presence or absence of 14 functional KIR genes (KIR2DL1, 2DL2, 2DL3, 2DL4, 2DL5, 3DL1, 3DL2, 3DL3, 2DS1, 2DS2, 2DS3, 2DS4, 2DS5, and 3DS1) using 2 PCR-SSP methods as previously described [27]. D/R KIR gene disparities were evaluated in both directions (D⁻R⁺ and D⁺R), in the HVG direction (D⁻R⁺ only), and in the GVH direction (D⁺R⁻ only).

The assignment of KIR genotypes was based on the reaction pattern observed compared to those associated with published KIR gene sequences. The KIR genotypes (AA, AB, and BB) were determined as initially defined by Uhrberg et al. [14].

Statistical Analysis 

Differences in categoric variables between 2 groups were evaluated by chi-square analysis (with Yates correction if needed or Fisher tests in case of limited series). Univariate and multivariate proportional hazard regression models were used to identify independent risk factors of death by means of log-rank tests and Cox proportional hazard models, respectively. The univariate Kaplan-Meier analysis was used to describe risk factors for death. Cumulative incidence using the competing risk method as described by Fine and Gray [31] was used to assess factors affecting the prognosis of aGVHD and relapse with death or death without relapse for relapse analysis as a competing event. Cox regression analysis was also used for the multivariate analysis of risk factors for death [32]. A stepwise backward procedure was used to construct a set of independent predictors of each end point. All predictors achieving a P-value below .15 were considered and sequentially removed if the P-value in the multiple model was above .05. The proportional hazard assumption was checked by graphical method. All tests were 2 sided, with the type I error rate fixed at .05. Statistical analyses were performed with SPSS 15 software, Stata 10 s and the R package “cmprsk” for competing risks. Numbers of cases at risk were given in parentheses in the Table 3, Table 4.

Table 2. HLA, KIR-Ligand, KIR, and KIR/KIR Ligand Gene Mismatching Characteristics of the 264 Unrelated Pairs Included in the Study

Variable	Categories	n (%)
D/R HLA identity∗
10/10 matched		164 (62.1)
not 10/10 matched		100 (37.9)
D/R KIR ligand mismatching status†
Bw4Bw6 / Bw6Bw6		2 (0.7)
Bw4Bw4 / Bw4Bw6		2 (0.7)
Bw6Bw6 / Bw4Bw6		2 (0.7)
C1C2 / C1C1		7 (2.6)
C1C2 / C2C2		9 (3.4)
C1C1 / C1C2		16 (6.0)
C2C2 / C1C2		4 (1.5)
Missing KIR ligand in patients‡
At least one KIR ligand absent:
Bw4 group absent		81 (30.6)
C1 group absent		40 (15.2)
C2 group absent		95 (36.0)
D/R KIR gene mismatching status§
2DL1	mismatched	23 (8.7)
2DL2	mismatched	124 (47.0)
2DL3	mismatched	48 (18)
2DL5	mismatched	77 (29.0)
3DL1	mismatched	36 (13.6)
2DS1	mismatched	121 (45.8)
2DS2	mismatched	126 (47.7)
2DS3	mismatched	101 (38.3)
2DS4	mismatched	33 (12.5)
2DS5	mismatched	96 (36.4)
3DS1	mismatched	127 (48.1)
D/R KIR genotype mismatching status¶
AA / AB		34 (12.9)
AB / AA		43 (16.3)
BB / AB		9 (3.4)
AB / BB		8 (3.0)
BB / AA		9 (3.4)
AA / BB		1 (0.4)
D inhibitory KIR/R KIR ligand matching status⊥
At least, one KIR/KIR ligand mismatch:
D 2DL1+/R C2−		89 (33.8)
D 2DL2+ and/or 2DL3+/R C1−		38 (14.3)
D 3DL1+/R Bw4−		73 (27.6)

Table 3. Significant Factors Influencing Acute GVHD Incidence, Overall Survival, and Relapse Incidence in Donor/Recipient Pairs of Unrelated HSCT

	All pairs (n=264)	P	HLA Identical Pairs (n=164)	P	HLA nonidentical pairs (n=100)	P
Outcome and significant factors, % (n)
Acute GVHD:∗
D/R KIR2DL5 mismatched versus matched			43% (26) versus 21% (93)	.01†‡
D/R KIR2DS1 mismatched versus matched					65% (49) versus 45% (51)	.02
D/R KIR2DS3 mismatched versus matched	14% (44) versus 28% (220)	.05†‡	12% (61) versus 29% (103)	.01†
D KIR genotype AA versus AB, BB			40% (15) versus 16% (76)	.03†§
At least 1 KIR/KIR ligand mismatch versus 0	24% (231) versus 41% (32)	.03†
D KIR2DL2−2DL3+2DS2−/R C1− versus R C1+	83% (12) versus 55% (99)	.04
Other factors
Myelogenous versus nonmyelogenous diseases	47% (147) versus 68% (80)	.007
D/R CMV+ versus other combinations	70% (38) versus 50% (208)	.01
Malignant versus nonmalignant diseases			56% (140) versus 29% (21)	.03
Overall survival:¶
Patients C2C2+ versus patients C1C1+, C1C2+	19% (21) versus 54% (126)	.005§
D/R KIR2DS1 mismatched versus matched			43% (47) versus 63% (44)	.03§
D/R KIR3DL1 mismatched versus matched	19% (36) versus 46% (228)	.004	23% (21) versus 53% (143)	.002	0% (9) versus 35% (91)	.01⊥
D/R KIR3DS1 mismatched versus matched			28% (37) versus 55% (126)	.03⊥
Other factors
D/R HLA nonidentical versus identical	31% (100) versus 49% (164)	.02
Grafts 1993-2000 versus 2001-2003	34% (177) versus 59% (87)	.001			25% (79) versus 61% (21)	.01
D/R CMV+ versus other combinations	23% (38) versus 46% (208)	.01	28% (31) versus 54% (152)	.0007
Grafts F/F versus other combinations	56% (61) versus 37% (203)	.03	74% (43) versus 42% (152)	.005
Relapse:∗∗
D/R KIR2DS3 mismatched versus matched			32% (31) versus 12% (109)	.02⊥
D/R KIR2DS4 mismatched versus matched	0% (27) versus 21% (202)	.05
D KIR3DL1+/ R Bw4− versus R Bw4+	28% (66) versus 18% (143)	.04
D KIR3DS1+/ R Bw4− versus R Bw4+	37% (26) versus 16% (72)	.02	33% (14) versus 8% (41)	.02
D KIR3DL1+, 3DS1+/ R Bw4− versus R Bw4+	43% (22) versus 20% (60)	.004	40% (11) versus 11% (33)	.03	44% (11) versus 29% (27)	.05

D indicates donor; R, recipient; F, female; CMV, cytomegalovirus; GVHD, graft-versus-host disease.

Table 4. Significant Factors Influencing Acute GVHD Incidence, Overall Survival, and Relapse Incidence in Donor/Recipient Pairs of Unrelated HSCT (Multivariate Analysis)

	All pairs (n=264)	P	HLA identical pairs (n=164)	P	HLA nonidentical pairs (n=100)	P
Outcome and significant factors, HR (95%CI)
Acute GVHD∗
D/R KIR2DL5 mismatched versus matched			2.63 (1.18-5.55)	.02†‡
D/R KIR2DS1 mismatched versus matched					1.88 (1.07-3.29)	.03
D/R KIR2DS3 mismatched versus matched			0.59 (0.37-0.95)	.03
D/R KIR2DS3 mismatched versus matched			0.25 (0.09-0.75)	.01†
Other factors
Myelogenous versus nonmyelogenous diseases	0.57 (0.38-0.84)	.004
D/R CMV+ versus other combinations	1.65 (1.02-2.68)	.04
Malignant versus nonmalignant diseases			2.67 (1.17-6.23)	.02
Overall survival§
Patient C2C2+ versus patients C1C1+, C1C2+	2.17 (1.26-3.75)	.005¶
D/R KIR3DL1 mismatched versus matched	1.94 (1.28-2.94)	.002	1.99 (1.14-3.51)	.02
Other factors
D/R HLA nonidentical versus identical	1.40 (1.00-1.95)	.05
D/R CMV+ versus other combinations	1.57 (1.03-2.41)	.04
Grafts F/F versus other combinations	0.58 (0.37-0.89)	.02	0.45 (0.23-0.89)	.02
Relapse⊥
D KIR3DL1+, 3DS1+/ R Bw4− versus R Bw4+	5.96 (2.13-16.67)	.001			4.72 (1.17-19.10)	.03

D indicates donor; R, recipient; F, female; CMV, cytomegalovirus; GVHD, graft-versus-host-disease.

Only factors P < .05 are included. Data are hazard ratio and 95% confidence interval.

Graft, patient, and donor characteristics—other than D/R HLA class I (KIR ligands) and/or KIR gene disparities—such as date of transplantation, pathologies (malignant versus nonmalignant diseases, myelogenous versus myelogenous diseases), conditioning regimen, D/R age, sex matching, and CMV serology that could potentially influence aGVHD occurrence, OS, relapse incidence, were also evaluated in all D/R pairs combined and then separately in HLA identical and nonidentical pairs. Those that were found significantly associated with aGVHD (pathologies and D/R cytomegalovirus (CMV) status) or with OS (D/R HLA class I and class II matching, D/R CMV status, and D/R sex combination, period of graft) by univariate analysis (Table 3) were included in the multivariate analysis with variables of interest for this study (Table 4). None of the previous characteristics was found significantly associated with relapse in our series by univariate method.

Back to Article Outline

Results 

Impact of Donor/Recipient KIR Ligand Disparities According to the “Ligand-Ligand” Model on Unrelated HSCT Outcome 

In this French cohort (n=264), 164 patients (62.1%) received transplants from fully 10/10 HLA matched Ds (ie, identical allele typing for HLA-A, -B, -Cw, -DRB1, and -DQB1 loci), whereas 100 (37.9%) received transplants from at least 1 HLA allele-mismatched D (Table 2). To evaluate HLA class I mismatching in terms of KIR-ligand mismatch status, D/R pairs were further divided into Bw4, Bw6, C1, or C2 groups depending on their HLA-A, -B, and -Cw typings. In the entire cohort (n=264), considering the Bw4 and Bw6 status, 6 D/R pairs (2.3%) were mismatched for the KIR ligand in both directions and only 2 D/R pairs were mismatched in the GVH direction (Table 2). In the entire cohort (n=264), considering the C1 and the C2 status, 36 D/R pairs (13.6%) were KIR ligand mismatched in both directions and 16 D/R pairs were KIR ligand mismatched in the GVH direction (Table 2). No significant differences were observed in the incidence of aGVHD, OS, and relapse incidence between KIR ligand matched versus KIR ligand mismatched D/R pairs in both directions and in the GVH direction only.

Impact of “Missing KIR Ligands” in Patients on Unrelated HSCT Outcome 

Taking into account both the Bw4 and C1 and C2 groups, 167 patients (63.3%) had at least 1 KIR ligand absent (Table 2). No significant differences were observed in the incidence of aGVHD, OS, and relapse incidence related to the absence of the HLA-Bw4 ligands in patients (data not shown). In contrast, the absence of HLA-Cw^Asp80 (C1) molecules in R had a significant impact on unrelated HSCT in this cohort because univariate analysis showed that C2C2⁺ (C1⁻, n=21) patients with myelogenous diseases had a lower OS compared to C1C1⁺ and C1C2⁺ recipients (C1⁺, n=126) (19% versus 54% at 2 years, P=.005; Table 3). Multivariate Cox analysis also revealed a decreased OS of C2C2⁺ (C1⁻) patients with myelogenous diseases (hazard ratio [HR]=2.17, P=.005; Table 4). A lack of the C1 ligand in R no longer had a significant effect on OS when HLA identical and nonidentical pairs were analyzed separately (data not shown). No significant differences were observed in terms of the incidence of aGVHD, or relapse related to the absence of the HLA-C1 ligands in patients (data not shown).

Impact of D/R KIR Gene Disparities According to the “Receptor-Receptor” Model on Unrelated HSCT Outcome 

Based on the presence or absence of individual KIR genes in each D/R pair, inhibitory and activating KIR gene disparities were evaluated (Table 2). The mismatching rate in both directions (R⁺D⁻ or D⁻R⁺) for KIR2DL1, 2DL2, 2DL3, 2DL5, and 3DL1 was 8.7%, 47%, 18%, 29%, and 13.6%, respectively. Concerning the activating KIR genes, the mismatching rate in both directions for KIR2DS1, 2DS2, 2DS3, 2DS4, 2DS5, and 3DS1 was 45.8%, 47.7%, 38.3%, 12.5%, 36.4%, and 48.1%, respectively. In terms of KIR genotypes, 104 D/R pairs (39%) were mismatched for AA, AB, or BB genotypes (Table 2). Univariate and multivariate analysis showed that mismatching for the KIR2DL5 gene only in the HVG direction (D⁻R⁺) was deleterious in HLA identical pairs by increasing severe aGVHD incidence (43% versus 21%, P=.01; Table 3; HR=2.63, P=.02; Table 4). D/R KIR2DS1 gene mismatches in both directions were deleterious in HLA identical pairs for patients with myelogenous diseases, by decreasing OS (43% versus 63%, P=.03; Table 3) and in HLA nonidentical pairs by increasing aGVHD incidence (65% versus 45%, P=.02; Table 3; HR=1.88, P=.03; Table 4). In contrast, D/R KIR2DS3 gene mismatches decreased severe aGVHD incidence both in all pairs combined (D⁻R⁺) and in HLA identical pairs (14% versus 28%, P=.05 and 12% versus 29%, P =.01 respectively; Table 3). This beneficial effect of D/R KIR2DS3 gene mismatches remained significant by multivariate analysis restricted to HLA identical pairs by decreasing both aGVHD_II-IV and severe aGVHD_III-IV (HR=0.59, P=.03 and HR=.25, P=.01 respectively; Table 4). Conversely, D/R KIR2DS3 mismatches in the GVH direction (D⁺R⁻) increased relapse incidence in HLA identical pairs restricted to malignant diseases (32% versus 12%, P=.02; Table 3). D/R KIR2DS4 gene mismatches in both directions had a beneficial effect on relapse incidence in all pairs restricted for malignant diseases (0% versus 21%, P=.05; Table 3), but this effect was no longer significant when D/R pairs were divided into HLA identical and nonidentical groups. Among these KIR genes, KIR3DL1 appeared as a “key marker” because D/R KIR3DL1 gene mismatches decreased OS in all pairs (19% versus 46%, P=.004; Table 3), in HLA identical pairs (23% versus 53%, P=.002; Table 3 and Figure 1A), as well as in HLA nonidentical pairs (0% versus 35%, P=.01, D⁺R^−; Table 3). Multivariate analysis also confirmed the deleterious effect of D/R KIR3DL1 gene mismatching on OS in all D/R pairs combined (HR=1.94, P=.002; Table 4) and in D/R HLA identical pairs (HR=1.99, P=.02; Table 4). Surprisingly, KIR3DS1, the activating counterpart of KIR3DL1, but which segregates as an allele, also had a deleterious effect, because HLA identical D/R pairs mismatched for KIR3DS1 in the GVH direction (D⁺R⁻) had a lower survival compared to KIR3DS1-matched D/R pairs (28% versus 55%, P=.03; Table 3 and Figure 1B). D/R KIR2DL1, 2DL2, 2DL3, 2DS2, and 2DS5 gene mismatches had no significant impact on unrelated HSCT outcome in this particular cohort (data not shown).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1 

  Impact of KIR3DL1/3DS1 gene disparities on patient survival in unrelated HSCT. (A) Cumulative survival of patients who received transplants matched for KIR3DL1 (n=143, R⁺D⁺ and R⁻D⁻) was compared with that of patients who received KIR3DL1 mismatched transplants in both directions (n=21, R⁺D⁻ and R⁻D⁺). Unrelated D/R pairs were identical at the allelic level for HLA-A, -B, -Cw, -DRB1, and -DQB1 (n=164). (B) Cumulative survival of patients who received transplants matched for KIR3DS1 (R⁺D⁺, R⁻D⁻) or mismatched in the HVG direction (R⁺D⁻) (n=126) was compared with that of patients who received KIR3DS1 mismatched transplants in the GVH direction (n=37, R⁻D⁺). Unrelated D/R pairs were identical at the allelic level for HLA-A, -B, -Cw, -DRB1, and -DQB1 (n=164). A value of P ≤ .05 was considered as statistically significant.

Patients with myelogenous diseases receiving grafts from AA KIR genotype donors (n=15) developed a higher incidence of severe aGVHD compared to patients receiving grafts from AB or BB KIR genotype donors restricted to D/R HLA identical pairs (40% versus 16%, P=.03; Table 3).

Impact of Donor KIR Genes and Recipient HLA Class I Ligand Disparities According to the “Receptor-Ligand” Model on Unrelated HSCT Outcome 

A potential correlation between the presence of inhibitory KIR2DL1, KIR2DL2/2DL3, or KIR3DL1 genes in the D and the absence of the corresponding KIR ligand in the R was next investigated. Overall, 200 D/R pairs had at least 1 KIR/KIR ligand-mismatch (Table 2). The severe aGVHD incidence was significantly decreased in D/R pairs with at least 1 KIR/KIR ligand mismatch compared to D/R pairs without KIR/KIR ligand mismatch (24% versus 41%, P=.03; Table 3). However, C1⁻ patients grafted with a KIR2DL2⁻/2DL3⁺/2DS2⁻ D showed a higher incidence of aGVHD (grades II-IV) compared to C1⁺ patients (83% versus 55%, P=.04; Table 3), suggesting the importance of some activating KIR genes in the donor. KIR3DL1/Bw4 or KIR3DL1/3DS1/Bw4 ligand mismatching had no significant effect on aGVHD incidence (data not shown). Although not significant, Bw4⁻ patients receiving grafts from KIR3DL1⁺/3DS1⁺ Ds displayed a lower OS both in HLA identical (37% versus 50%, P=.56) and in HLA nonidentical pairs (19% versus 32%, P=.16, data not shown). Interestingly, the deleterious effect of the donor KIR3DL1⁺ or KIR3DS1⁺ gene was relevant for relapse incidence in Bw4⁻ patients with malignant diseases. Indeed, Bw4⁻ patients receiving grafts from KIR3DL1⁺ Ds or from KIR3DS1⁺ Ds showed a higher incidence of relapse compared to Bw4⁺ patients (28% versus 18%, P=.04 and 37% versus 16%, P=.02, respectively; Table 3) when all D/R pairs were taken into account. When restricted to HLA identical D/R pairs, only Bw4⁻ patients with malignant diseases receiving a graft from KIR3DS1⁺ donors exhibited a significantly higher incidence of relapse compared to Bw4⁺ patients (33% versus 8%, P=.02; Table 3 and Figure 2A). This effect was reinforced by taking into account both the presence of KIR3DL1 and the KIR3DS1 gene in the D, because the incidence of relapse increased to 43% for Bw4⁻ patients receiving grafts from KIR3DL1⁺/3DS1⁺ Ds compared to 20% for Bw4⁺ patients when all D/R pairs were considered (P=.004; Table 3; HR=5.96, P=.001; Table 4). This deleterious combination (D: KIR3DL1⁺/3DS1⁺ and R Bw4⁻) had a significant impact both in HLA identical pairs (40% versus 11%, P=.03; Table 3 and Figure 2B) and in HLA nonidentical pairs (44% versus 29%, P=.05; Table 3 and Figure 2C; HR=4.72, P=.03; Table 4).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2 

  Impact of donor KIR3DL1/3DS1 and recipient Bw4 ligand disparities on relapse incidence in unrelated HSCT. (A) Cumulative incidence of relapse in HLA-Bw4⁺ patients with malignant diseases grafted with a KIR3DS1⁺ donor (n=41) was compared to that in HLA-Bw4⁻ patients receiving a graft from a KIR3DS1⁺ donor (n=14). Unrelated D/R pairs were HLA identical at the allelic level for HLA-A, -B, -Cw, -DRB1, and -DQB1. (B) Cumulative incidence of relapse in HLA-Bw4⁺ patients with malignant diseases receiving a graft from a KIR3DL1⁺/3DS1⁺ donor (n=33) was compared to that in HLA-Bw4⁻ patients receiving a graft from a KIR3DL1⁺/3DS1⁺ donor (n=11). Unrelated D/R pairs were HLA identical at allelic level for HLA-A, -B, -Cw, -DRB1, and -DQB1. (C) Cumulative incidence of relapse in HLA-Bw4⁺ patients with malignant diseases receiving a graft from a KIR3DL1⁺/3DS1⁺ donor (n=27) was compared to that in HLA-Bw4⁻ patients receiving a graft from a KIR3DL1⁺/3DS1⁺ donor (n=11). Unrelated D/R pairs were HLA mismatched with at least 1 allele mismatch for HLA-A, -B, -Cw, -DRB1, or -DQB1 loci. A value of p ≤ .05 was considered as statistically significant. The Bw4 status was evaluated depending on HLA-A (A23, A24, A32) and -B allelic typing of the recipient.

Back to Article Outline

Discussion 

In this retrospective, multicentric study, we analyzed the clinical impact of the 4 defined NK-alloreactivity models in 264 patients who underwent T repleted unrelated HSCT. First, we did not observe that the “KIR ligand-ligand” model had a significant clinical impact on unrelated HSCT outcome. However, the effect of KIR ligand mismatching (Bw4, C1, or C2) in both directions and in the GVH direction (D⁺R⁻) was evaluated in few KIR-ligand mismatched pairs compared to the KIR ligand matched pairs. This observation is probably also related to the absence of T cell depletion in the grafts or to the heterogeneous diseases included in this study, because the beneficial effect of KIR ligand disparities in the GVH direction was initially described in haploidentical HSCT with extensive T cell depletion and was restricted to acute myelogenous leukemia (AML) patients [3]. Second, we observed that the “missing KIR ligand” model had a significant but limited effect on unrelated HSCT. Indeed, only the absence of C1 ligand in patients was associated with a decreased OS when HLA identical and HLA nonidentical D/R pairs were combined. However, a lack of C1 ligand in patients no longer had a significant effect on OS when HLA identical and nonidentical pairs were analysed separately or when the presence of the D KIR2DS2 gene was taken into account as initially reported in HLA-identical sibling HSCT for myelogenous leukemia [33]. Moreover, no effect of missing KIR ligand on aGVHD or relapse incidence was observed, in contrast to previous studies 20, 34. Third, we observed that the “KIR receptor-receptor” model influenced the outcome of both HLA identical and HLA nonidentical unrelated HSCT. In particular, multivariate analyses showed that D/R KIR2DL5 or KIR2DS1 mismatches are deleterious by increasing aGVHD incidence in HLA identical and HLA nonidentical pairs, respectively. In contrast, D/R KIR2DS3 mismatches had a beneficial effect by decreasing aGVHD in HLA identical pairs. The D KIR3DL1 gene appeared to be a risk factor, especially in HLA identical D/R pairs, because mismatches in the GVH direction (D⁺R⁻) significantly decreased OS. This deleterious effect of the donor KIR3DL1 gene was reinforced taking into account the presence of its activating counterpart (ie, KIR3DS1). Our present findings concerning the deleterious effects of the D/R KIR gene mismatches are consistent with the literature because inhibitory and mainly activating KIR gene disparities in the GVH direction (D⁺R⁻) have been correlated with a higher incidence of aGVHD, a higher incidence of relapse, or a lower OS in unrelated HSCT pairs 21, 35, 36, 37, 38. The presence of particular activating KIR genes in the D (ie, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS5, KIR3DS1) seems deleterious both in Caucasoid 21, 35, 36, 37, 38 and in non-Caucasoid populations [39]. Two recent studies have reported a deleterious effect of the D KIR3DS1 gene [37] or the R KIR2DS3 gene [40] on relapse incidence in unrelated HSCT. Surprisingly, in terms of KIR haplotypes, the presence of a high number of activating KIR genes in the D (AB and BB vs. AA haplotypes) had no significant effect on OS or relapse incidence, but decreased aGVHD in our cohort. This beneficial effect of D KIR B haplotypes has also been demonstrated in unrelated HSCT outcome in Korean patients, by decreasing aGVHD incidence [41], and recently in a large cohort of AML patients, by increasing OS [42].

Finally, we observed that the “KIR receptor-ligand” model influences the outcome of both HLA identical and nonidentical D/R unrelated pairs. In particular, we showed for the first time that the genetic combination of D KIR3DL1⁺/3DS1⁺ and R Bw4⁻ correlated with the highest incidence of relapse in both HLA identical and HLA nonidentical D/R unrelated pairs compared to the combination D KIR3DL1⁺/3DS1⁺ and R Bw4⁺. Until now, few studies have investigated the impact of the “KIR receptor-ligand” model on unrelated HSCT outcome and no relevant combination between the presence of particular donor KIR genes and the absence of a corresponding KIR ligand in the R has been identified 37, 39, 42, 43, 44. However, the studies published so far have focused only on limited KIR/KIR ligand interactions (mainly KIR2DL1/C2, KIR2DL2/2DL3/C1), or limited KIR ligands (only HLA-B Bw4⁺ molecules considered as KIR3DL1 ligands). In this study, we demonstrated for the first time that the presence of the D KIR3DL1⁺/3DS1⁺ gene and the absence of the corresponding R Bw4 ligand (both HLA-A and HLA-B molecules considered) correlates with a higher incidence of relapse, whereas the presence of the D KIR3DL1⁺/3DS1⁺ gene and the R Bw4 ligand correlates with a lower incidence of relapse. These immunogenetic data suggest a potential beneficial D KIR3DL1⁺/3DS1⁺ NK cell alloreactivity against R Bw4+ leukemic cells, leading to a decreased incidence of relapse (potential GVL effect), as already suggested by Foley et al. [45]. However, the expression level of the KIR3DL1 and KIR3DS1 receptors at the surface of D NK cells should be measured because the D KIR3DL1⁺/3DS1⁺ NK repertoire could differ depending on the presence or absence of autologous KIR ligand (Bw4) and on KIR3DL1 allelic and/or promoter polymorphism 46, 47, 48.

Overall, our results show that the “KIR receptor-receptor” and “KIR receptor-ligand” models seemed the most suitable to predict NK alloreactivity, because they had a significant impact on aGVHD occurrence, OS, and relapse incidence in HLA identical and in HLA nonidentical unrelated D/R pairs. In particular, our results suggest a detrimental effect of D KIR3DL1⁺/3DS1⁺ D NK cells transplanted into HLA-Bw4⁻ patients in the absence of an education process via KIR3DL1/HLA-Bw4 interactions. These results have to be confirmed on independent homogenous series taking into account other important parameters that may also contribute to HSCT outcome such as GVHD prophylaxis regimen, disease status, or high- versus low-risk hematologic malignancies.

Back to Article Outline

Acknowledgments 

The authors thank the data manager and the transplant physicians of the SFGM-TC (Société Française de Greffe de Moelle et de Thérapie Cellulaire) for their collaboration, and Dr. Christelle Retière (EA4271, EFS Nantes) for critically reading the manuscript.

Financial disclosure: This work was supported by grants from the “Fondation pour la Recherche Médicale ARS 2000,” “France Greffe de Moelle 2005,” and “IRGHET 2007.”

Back to Article Outline

References 

1. Hansen JA, Petersdorf EW, Lin MT, et al. Genetics of allogeneic hematopoietic cell transplantation. Role of HLA matching, functional variation in immune response genes. Immunol Res. 2008;41:56–78
   • View In Article
   • CrossRef
2. Barrett AJ. Understanding and harnessing the graft-versus-leukaemia effect. Br J Haematol. 2008;142:877–888
   • View In Article
   • CrossRef
3. Ruggeri L, Capanni M, Urbani E, et al. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science. 2002;295:2097–2100
   • View In Article
   • CrossRef
4. Vivier E, Tomasello E, Baratin M, Walzer T, Ugolini S. Functions of natural killer cells. Nat Immunol. 2008;9:503–510
   • View In Article
   • CrossRef
5. Parham PMHC. class I molecules and KIRs in human history, health and survival. Nat Rev Immunol. 2005;5:201–214
   • View In Article
   • MEDLINE
6. Stern M, Ruggeri L, Capanni M, Mancusi A, Velardi A. Human leukocyte antigens A23, A24, and A32 but not A25 are ligands for KIR3DL1. Blood. 2008;112:708–710
   • View In Article
   • CrossRef
7. Stewart CA, Laugier-Anfossi F, Vely F, et al. Recognition of peptide-MHC class I complexes by activating killer immunoglobulin-like receptors. Proc Natl Acad Sci USA. 2005;102:13224–13229
   • View In Article
8. Chewning JH, Gudme CN, Hsu KC, Selvakumar A, Dupont B. KIR2DS1-positive NK cells mediate alloresponse against the C2 HLA-KIR ligand group in vitro. J Immunol. 2007;179:854–868
   • View In Article
9. Morvan M, David G, Sebille V, et al. HLA KIR ligand environments and activating KIR control KIR NK-cell functions. Eur J Immunol. 2008;38:3474–3486
   • View In Article
10. O'Connor GM, Guinan KJ, Cunningham RT, Middleton D, Parham P, Gardiner CM. Functional polymorphism of the KIR3DL1/S1 receptor on human NK cells. J Immunol. 2007;178:235–241
    • View In Article
    • MEDLINE
11. Gillespie GM, Bashirova A, Dong T, McVicar DW, Rowland-Jones SL, Carrington M. Lack of KIR3DS1 binding to MHC class I Bw4 tetramers in complex with CD8 + T cell epitopes. AIDS Res Hum Retroviruses. 2007;23:451–455
    • View In Article
    • MEDLINE
    • CrossRef
12. Middleton D, Williams F, Halfpenny IA. KIR genes. Transpl Immunol. 2005;14:135–142
    • View In Article
    • MEDLINE
    • CrossRef
13. Single RM, Martin MP, Meyer D, Gao X, Carrington M. Methods for assessing gene content diversity of KIR with examples from a global set of populations. Immunogenetics. 2008;60:711–725
    • View In Article
    • CrossRef
14. Uhrberg M, Valiante NM, Shum BP, et al. Human diversity in killer cell inhibitory receptor genes. Immunity. 1997;7:753–763
    • View In Article
    • MEDLINE
    • CrossRef
15. Robinson J, Waller MJ, Fail SC, Marsh SG. The IMGT/HLA and IPD databases. Hum Mutat. 2006;27:1192–1199
    • View In Article
    • CrossRef
16. Ruggeri L, Mancusi A, Burchielli E, et al. NK cell alloreactivity and allogeneic hematopoietic stem cell transplantation. Blood Cells Mol Dis. 2008;40:84–90
    • View In Article
    • CrossRef
17. Willemze R, Rodrigues CA, Labopin M, et al. KIR-ligand incompatibility in the graft-versus-host direction improves outcomes after umbilical cord blood transplantation for acute leukemia. Leukemia. 2009;23:492–500
    • View In Article
    • CrossRef
18. Bignon JD, Gagne K. KIR matching in hematopoietic stem cell transplantation. Curr Opin Immunol. 2005;17:553–559
    • View In Article
    • MEDLINE
    • CrossRef
19. Witt CS, Christiansen FT. The relevance of natural killer cell human leucocyte antigen epitopes and killer cell immunoglobulin-like receptors in bone marrow transplantation. Vox Sang. 2006;90:10–20
    • View In Article
    • MEDLINE
    • CrossRef
20. Hsu KC, Keever-Taylor CA, Wilton A, et al. Improved outcome in HLA-identical sibling hematopoietic stem-cell transplantation for acute myelogenous leukemia predicted by KIR and HLA genotypes. Blood. 2005;105:4878–4884
    • View In Article
    • MEDLINE
    • CrossRef
21. Gagne K, Brizard G, Gueglio B, et al. Relevance of KIR gene polymorphisms in bone marrow transplantation outcome. Hum Immunol. 2002;63:271–280
    • View In Article
    • MEDLINE
    • CrossRef
22. Leung W, Iyengar R, Turner V, et al. Determinants of antileukemia effects of allogeneic NK cells. J Immunol. 2004;172:644–650
    • View In Article
    • MEDLINE
23. Shilling HG, Young N, Guethlein LA, et al. Genetic control of human NK cell repertoire. J Immunol. 2002;169:239–247
    • View In Article
    • MEDLINE
24. Yokoyama WM, Kim S. Licensing of natural killer cells by self-major histocompatibility complex class I. Immunol Rev. 2006;214:143–154
    • View In Article
    • MEDLINE
    • CrossRef
25. Yawata M, Yawata N, Draghi M, Partheniou F, Little AM, Parham P. MHC class I-specific inhibitory receptors and their ligands structure diverse human NK-cell repertoires toward a balance of missing self-response. Blood. 2008;112:2369–2380
    • View In Article
    • CrossRef
26. Kim S, Sunwoo JB, Yang L, et al. HLA alleles determine differences in human natural killer cell responsiveness and potency. Proc Natl Acad Sci USA. 2008;105:3053–3058
    • View In Article
27. Gagne K, Busson M, Balere-Appert ML, et al. Relevance of KIR gene matching in unrelated hematopoietic stem cell transplantations. Tissue Antigens. 2007;69(Suppl 1):118–122
    • View In Article
    • CrossRef
28. Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HLA-matched sibling donors. Transplantation. 1974;18:295–304
    • View In Article
    • MEDLINE
    • CrossRef
29. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15:825–828
    • View In Article
    • MEDLINE
30. Loiseau P, Busson M, Balere ML, et al. HLA Association with hematopoietic stem cell transplantation outcome: the number of mismatches at HLA-A, -B, -C, -DRB1, or -DQB1 is strongly associated with overall survival. Biol Blood Marrow Transplant. 2007;13:965–974
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (226 KB)
    • CrossRef
31. Fine JP, Gray RJ. A proportional hazards model for the sub distribution of a competing risk. JASA. 1999;94:496
    • View In Article
32. Cox DR. Regression models and life tables (with discussion). J R Stat Soc. 1972;34:187–220
    • View In Article
33. Cook MA, Milligan DW, Fegan CD, et al. The impact of donor KIR and patient HLA-C genotypes on outcome following HLA-identical sibling hematopoietic stem cell transplantation for myeloid leukemia. Blood. 2004;103:1521–1526
    • View In Article
    • MEDLINE
    • CrossRef
34. Miller JS, Cooley S, Parham P, et al. Missing KIR ligands are associated with less relapse and increased graft-versus-host disease (GVHD) following unrelated donor allogeneic HCT. Blood. 2007;109:5058–5061
    • View In Article
    • MEDLINE
    • CrossRef
35. Sun JY, Gaidulis L, Dagis A, et al. Killer Ig-like receptor (KIR) compatibility plays a role in the prevalence of acute GVHD in unrelated hematopoietic cell transplants for AML. Bone Marrow Transplant. 2005;36:525–530
    • View In Article
    • MEDLINE
    • CrossRef
36. De Santis D, Bishara A, Witt CS, et al. Natural killer cell HLA-C epitopes and killer cell immunoglobulin-like receptors both influence outcome of mismatched unrelated donor bone marrow transplants. Tissue Antigens. 2005;65:519–528
    • View In Article
    • MEDLINE
    • CrossRef
37. Kroger N, Binder T, Zabelina T, et al. Low number of donor activating killer immunoglobulin-like receptors (KIR) genes but not KIR-ligand mismatch prevents relapse and improves disease-free survival in leukemia patients after in vivo T-cell depleted unrelated stem cell transplantation. Transplantation. 2006;82:1024–1030
    • View In Article
    • MEDLINE
    • CrossRef
38. Giebel S, Locatelli F, Lamparelli T, et al. Survival advantage with KIR ligand incompatibility in hematopoietic stem cell transplantation from unrelated donors. Blood. 2003;102:814–819
    • View In Article
    • MEDLINE
    • CrossRef
39. Yabe T, Matsuo K, Hirayasu K, et al. Donor killer immunoglobulin-like receptor (KIR) genotype-patient cognate KIR ligand combination and antithymocyte globulin preadministration are critical factors in outcome of HLA-C-KIR ligand-mismatched T cell-replete unrelated bone marrow transplantation. Biol Blood Marrow Transplant. 2008;14:75–87
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (866 KB)
    • CrossRef
40. Schellekens J, Rozemuller EH, Petersen EJ, van den Tweel JG, Verdonck LF, Tilanus MG. Patients benefit from the addition of KIR repertoire data to the donor selection procedure for unrelated haematopoietic stem cell transplantation. Mol Immunol. 2008;45:981–989
    • View In Article
    • CrossRef
41. Kim SY, Choi HB, Yoon HY, et al. Influence of killer cell immunoglobulin-like receptor genotypes on acute graft-vs-host disease after unrelated hematopoietic stem cell transplantation in Koreans. Tissue Antigens. 2007;69(Suppl 1):114–117
    • View In Article
    • CrossRef
42. Cooley S, Trachtenberg E, Bergemann TL, et al. Donors with group B KIR haplotypes improve relapse-free survival after unrelated hematopoietic cell transplantation for acute myelogenous leukemia. Blood. 2009;113:726–732
    • View In Article
    • CrossRef
43. La Nasa G, Littera R, Locatelli F, et al. Status of donor-recipient HLA class I ligands and not the KIR genotype is predictive for the outcome of unrelated hematopoietic stem cell transplantation in beta-thalassemia patients. Biol Blood Marrow Transplant. 2007;13:1358–1368
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (375 KB)
    • CrossRef
44. Ludajic K, Balavarca Y, Bickeboller H, et al. KIR genes and KIR ligands affect occurrence of acute GVHD after unrelated, 12/12 HLA matched, hematopoietic stem cell transplantation. Bone Marrow Transplant. 2009;44:97–103
    • View In Article
    • CrossRef
45. Foley BA, De Santis D, Van Beelen E, Lathbury LJ, Christiansen FT, Witt CS. The reactivity of Bw4 + HLA-B and HLA-A alleles with KIR3DL1: implications for patient and donor suitability for haploidentical stem cell transplantations. Blood. 2008;112:435–443
    • View In Article
    • CrossRef
46. Gardiner CM, Guethlein LA, Shilling HG, et al. Different NK cell surface phenotypes defined by the DX9 antibody are due to KIR3DL1 gene polymorphism. J Immunol. 2001;166:2992–3001
    • View In Article
    • MEDLINE
47. Carr WH, Pando MJ, Parham P. KIR3DL1 polymorphisms that affect NK cell inhibition by HLA-Bw4 ligand. J Immunol. 2005;175:5222–5229
    • View In Article
    • MEDLINE
48. Morvan M, Willem C, Gagne K, et al. Phenotypic and functional analyses of KIR3DL1+ and KIR3DS1 + NK cell subsets demonstrate differential regulation by Bw4 molecules and induced KIR3DS1 expression on stimulated NK cells. J Immunol. 2009;182:6727–6735
    • View In Article
    • CrossRef