Although allogeneic hematopoietic stem-cell transplantation (HSCT) provides high cure rates for children with high-risk acute lymphoblastic leukaemia (ALL), relapses remain the main cause of treatment failure. Whereas donor killer cell immunoglobulin-like receptor (KIR) genotype was shown to impact on relapse incidence in adult myeloid leukaemia similar studies in paediatric ALL are largely missing. Effect of donor KIR genotype on transplant outcome was evaluated in 317 children receiving a first myeloablative HSCT from an HLA-matched unrelated donor or sibling within the prospective ALL-SCT-BFM-2003 trial. Analysis of donor KIR gene polymorphism revealed that centromeric presence and telomeric absence of KIR B haplotypes was associated with reduced relapse risk. A centromeric/telomeric KIR score (ct-KIR score) integrating these observations correlated with relapse risk (hazard ratio (HR) 0.58; P = 0.002) while it had no impact on graft-versus-host disease or non-relapse mortality. In multivariable analyses ct-KIR score was associated with reduced relapse risk (HR 0.58; P = 0.003) and a trend towards improved event-free survival (HR 0.76; P = 0.059). This effect proved independent of MRD level prior to HSCT. Our data suggest that in children with ALL undergoing HSCT after myeloablative conditioning, donor selection based on KIR genotyping holds promise to improve clinical outcome by decreasing relapse risk and prolonged event-free survival.
Allogeneic hematopoietic stem-cell transplantation (HSCT) provides high cure rates in children suffering from very high-risk primary or relapsed acute lymphoblastic leukaemia (ALL) who display poor prognosis with conventional chemotherapy. Recent improvements in donor selection relying on high-resolution human leucocyte antigen (HLA) typing as well as optimisation of supportive care, including effective graft-versus-host-disease (GvHD) prophylaxis and early diagnosis and pre-emptive treatment of infections, have substantially reduced non-relapse morbidity and mortality, in particular in the setting of transplantation from HLA-matched unrelated donors (MUD). Consequently, current results of HSCT from MUDs are comparable to those from HLA-identical sibling donors (ISD) and approach 4-year event-free survival rates of 65–70% . Relapses of leukaemia, however, represent the main cause of treatment failure, thus warranting investigation of novel strategies to improve leukaemia control in paediatric HSCT.
As therapeutic efficacy of HSCT in addition to high dose chemo-/radiotherapy critically relies on immune effects provided by allogeneic donor cells various strategies to identify the immunologically optimal stem-cell donor have been promoted. Among those, the employment of leukaemia-directed effects provided by donor natural killer (NK) cells play a prominent role . The function of NK cells is regulated by killer cell immunoglobulin-like receptors (KIR) that comprise of 14 polymorphic inhibitory or stimulatory receptors whose genes vary in number and content between individuals. The specific set of KIR genes present in an individual at the independently segregating centromeric and telomeric KIR gene loci defines two common haplotypes: Group A haplotypes provide high-affinity recognition of all major HLA class I-encoded ligands to secure self-tolerance, whereas group B haplotypes have variable gene content contributing to the diversification of NK cell repertoires .
Recent large registry-based analyses primarily focusing on adult patients suffering from acute myeloblastic leukaemia (AML) have identified donor selection based on presence and chromosomal position of the KIR B haplotype, with a predominant role of the centromeric KIR genes KIR2DL2/2DS2, to provide superior leukaemia control [4,5,6]. Subsequent studies, however, have revealed that the effect of donor KIR haplotypes is highly context-dependent, i.e., depending on the specific transplant scenario with regard to underlying malignancy, conditioning intensity, donor type, and in vivo and/or in vitro T cell depletion. [5, 7,8,9].
As data in the paediatric HLA-matched setting are largely missing, we here evaluated the impact of donor KIR haplotypes and their respective chromosomal position on the clinical outcome of 317 children with ALL homogeneously treated within the ALL-SCT-BFM-2003 trial.
Patients and donors
This study retrospectively investigated the impact of donor KIR genotype on outcome in paediatric patients who received a first myeloablative HSCT from an HLA-identical sibling or an HLA-matched unrelated donor within the prospective ALL-SCT-BFM-2003 trial. Between September 2003 and September 2011, 411 children with high-risk ALL, who received allogeneic SCT, were enroled in the ALL-BFM-SCT 2003 trial . From 317 of these 411 patients and their stem-cell donors DNA samples could be collected and KIR-typed (Table S1). Ninety-four patients were not included in this study as DNA samples from donors were not available. Informed consent was obtained in accordance with the Declaration of Helsinki.
Kir typing and Kir haplotype assignment
KIR genotyping of 12 KIR genes (KIR2DL1, KIR2DL2, KIR2DL3, KIR2DL5, KIR3DL1, KIR3DL2, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS4, KIR2DS5, and KIR3DS1) was performed on genomic DNA from related and unrelated donors of hematopoietic stem-cells by polymerase chain reaction with sequence specific primers (PCR-SSP), as previously described . A total of 10% randomly selected samples were retyped by an independent KIR typing method with full concordance . Based on the presence/absence polymorphism and the strong negative linkage disequilibrium of some KIR genes, the centromeric and telomeric modules that constitute a KIR haplotype can be defined as follows: presence of KIR2DL1 and KIR2DL3 denotes a centromeric A (Cen-A) module whereas presence of KIR2DL2 and/or KIR2DS2 denotes a centromeric B (Cen-B) module. On the telomeric part of the KIR locus the presence of KIR3DL1 and KIR2DS4 indicates the telomeric A (Tel-A) module while presence of KIR3DS1 and/or KIR2DS1 indicates the telomeric B (Tel-B) module.
For non-time to event variables, χ2 test was used to compare groups for categorical variables. Wilcoxon rank-sum test (Kruskal–Wallis test for more than two populations) was used for continuous variables. The overall and the event-free survival probabilities were evaluated using the Kaplan–Meier method and the log-rank test. Cumulative incidences of events were calculated by the method of Kalbfleisch and Prentice , and compared using the Gray test . Competing events were defined as follows: for relapse incidence: deaths from any cause and secondary malignancies; for chronic GVHD: death from any cause; and for non-relapse mortality: relapses and secondary malignancies. The impact of donor KIR haplotypes and the donor ct-KIR score on time-dependent outcome variables (cumulative incidence of relapse, chronic GvHD and non-relapse mortality; probability of event-free and overall survival) was further analysed in univariable und multivariate Cox regression models.
The Cox proportional hazards model was used to estimate survival, and the proportional subdistribution hazards model of Fine and Gray  for censored data subject to competing risks was applied for analyses of relapse incidence and non-relapse mortality. In order to perform a trend test the ct-KIR score was included into the models as a numerical variable. More information on statistical methods is provided in the online supplement.
Patient and donor characteristics
Characteristics of the study patients and their donors are listed in Table 1. After conditioning, all patients received stem-cell transplantations during first or later remission standardised with regard to donor selection, myeloablative conditioning, donor type-stratified GvHD prophylaxis and supportive care. Depending on donor availability, grafts originated from HLA-identical siblings (n = 86) or from ≥9/10 HLA-matched unrelated donors (n = 231). As shown in Table S1, a lower proportion of patients transplanted in >CR2 was the sole patient, disease and donor characteristic with significant difference between our study cohort and the 94 patients not included in the ALL-BFM-SCT 2003 trial.
Effect of presence and absence of donor centromeric and telomeric KIR B motifs
As previously reported by Cooley et al. and Bachanova et al., a relapse protective effect of donor KIR B haplotype motifs was identified in adult AML and non-Hodgkin lymphoma patients [4,5,6, 15]. In particular, centromeric group B motifs were found to be associated with improved disease control in these studies. To investigate the role of KIR B motifs in children with ALL in a candidate approach, patients were grouped according to the presence of donor B motifs in the centromeric (CenB+ vs. CenB−) and the telomeric region (TelB+ vs. TelB−). In accordance with the studies by Cooley and Bachanova, donors lacking a centromeric KIR B motif (CenB−) were associated with significantly increased relapse incidence (RI) (26% 95%-CI 23–29%) compared to donors with at least one centromeric KIR B motif (CenB+; 17% 95%-CI 14–20%; hazard ratio, 1.64; 95%-CI, 1.00–2.70; P = 0.048; Fig. 1a). In addition, we found that patients transplanted from donors with at least one telomeric KIR B motif (TelB+) experienced a higher relapse incidence (RI 27% 95%-CI 23–31%) compared to donors without telomeric KIR B motifs (TelB-; RI 17% 95%-CI 14–20%; hazard ratio, 1.65; 95%-CI, 1.03–2.65; P = 0.039; Fig. 1b). Together, these findings suggest that the absence of centromeric and presence of telomeric group B haplotypes are independent risk factors for relapse in childhood ALL.
To evaluate an independent effect of centromeric KIR B motifs vs telomeric KIR B motifs donors were divided into CenB−TelB−, CenB+TelB−, CenB−TelB+ and CenB+TelB+ groups. CenB + TelB- donors were associated with the lowest RI (12% 95%-CI 8–16%) while patients transplanted from CenB-TelB+donors fared significantly worse (34% 95%-CI 28–40%). Patients who were transplanted from either CenB + TelB + or CenB−TelB− donors experienced intermediate relapse incidences (22% 95%-CI 17–27% and 21% 95%-CI 17–25%, respectively; Fig. 1c). These data confirm an independent effect of centromeric and telomeric KIR B motifs on relapse incidence.
To harness the dichotomous role of KIR B motifs in the centromeric vs. telomeric gene locus we integrated the individual donor KIR genotypes at these two regions into a composite centromeric/telomeric KIR (ct-KIR) score: The ct-KIR score was assigned the value 0 in donors with KIR haplotypes associated with higher relapse risk at both centromeric and telomeric position (i.e. CenB- and TelB+), the value 1 in donors with the presence of one KIR haplotype associated with higher relapse risk at either centromeric or telomeric position (CenB+ and TelB+ or CenB− and TelB−), and the value 2 in donors lacking any centromeric and telomeric KIR haplotype associated with higher relapse risk (CenB+ and TelB−; Table 2). As shown in Fig. 1d, the donor ct-KIR score clearly discriminated paediatric ALL transplant recipients with high (ct-KIR score 0, n = 59), moderate (ct-KIR score 1, n = 175) and low (ct-KIR score 2, n = 83) cumulative incidences of relapse of 34% 95%-CI 28–40% vs. 21% 95%-CI 18–24% vs. 12% 95%-CI 8–16% (hazard ratio, 0.58; 95%-CI, 0.41–0.82; P = 0.002). Thus, with regard to relapse risk the ct-KIR score identifies unfavourable (ct-KIR score 0), intermediate (ct-KIR score 1) and favourable (ct-KIR score 2) donors (Fig. 1d) and this holds true for transplantations from both HLA-identical sibling donors and HLA-matched unrelated donors (Fig. S1 and S2).
Accounting for the current unavailability of an external validation cohort for this potentially practice-changing finding, we employed the method of bootstrapping to internally validate our ct-KIR score definition procedure. As described above, the ct-KIR score was defined according to the presence or absence of donor B motifs in the way to ensure an optimal association between the donor KIR status and the cumulative relapse incidence within the cohort of 317 patients. A potential statistical problem (model overfitting) related to the circumstance that we used the complete dataset to build the score, is that the model performance ability may shrink if measured on a new data set.
Our validation approach allowed for calculating the effect of model overfitting on the cumulative incidence of relapse for the subgroups of patients identified by the ct-KIR score: 33% (20-44%) for ct-KIR score 0, 22% (17-29%) for ct-KIR score 1, and 14% (8-22%) for ct-KIR score 0 (hazard ratio, 0.63; 95%-CI, 0.46-0.87; P = 0.005). These results validate the robust predictive value of the donor ct-KIR score on the cumulative incidence of relapse (Fig. 1e).
Having identified this significant association of donor KIR status with relapse incidence we strived for carefully evaluating potential biases due to unequally distributed confounders between the patient groups discriminated by donor ct-KIR score. Donor KIR genotype and thus ct-KIR score status is segregating independently from other donor characteristics like HLA genotype and AB0 blood group based on their genomic positioning on different chromosomes. As shown in Table 1 donor ct-KIR score did not correlate with any patient, disease or donor characteristic except for a slightly higher prevalence of HLA-identical sibling donors in the unfavourable donor group. Importantly, patients with high minimal residual disease (MRD) burden prior to HSCT, representing one major risk factor for post-transplant relapses (Fig. S3), were equally distributed among the donor groups (Table 1). Taken together, these observations illustrate random distribution of the ct-KIR score in donors of our study cohort.
In order to confirm or reject an independent prognostic impact of donor ct-KIR score on relapse incidence, we performed multivariable analysis adjusting for patient, disease and donor selection criteria such as immunophenotype, disease recurrence risk (Fig. S4 and S5), donor type [16,17,18,19], CMV serostatus , sex match and AB0 blood group . As shown in Table 3 donor ct-KIR score was independently associated with reduced relapse incidence (hazard ratio for one-point increment in ct-KIR score, 0.58; 95%-CI, 0.40–0.83; P = 0.003). The other factor reaching statistical significance in multivariable analysis was disease recurrence risk (hazard ratio, 2.84; 95%-CI, 1.64–4.92; P < 0.001) whereas no other patient, disease or donor characteristic including donor KIR – recipient KIR ligand mismatch had any effect on relapse incidence (Table 3 and Fig. S6). Given the fact that MRD level prior to HSCT has recently emerged as a prognostic factor for relapse risk , we performed additional multivariable analyses within all patients with available MRD data revealing that ct-KIR score and MRD level were the only independent risk factors for relapse thus validating that the effect of the donor ct-KIR score acts independently from MRD level prior to transplant (Table S3). Finally, donor ct-KIR score proved of prognostic significance within the subgroup of patients with low risk MRD level, while the subgroup of patient with high risk MRD was too small to study (Fig. S7).
Effect of donor Ct-KIR score on acute GvHD, chronic GvHD and non-relapse mortality
We next explored whether the relapse protective impact of donor ct-KIR status was achieved at the expense of increased risk of graft-versus-host-disease (GvHD). As shown in Table 4 for acute GvHD and Fig. 2a, b for any grade and extensive chronic GvHD we could not identify any detectable impact of donor ct-KIR status on GvHD risk. Most importantly, the incidence of extensive chronic GvHD representing the major cause of morbidity and non-relapse mortality in paediatric allogeneic HSCT  proved low and identical in all three donor ct-KIR score groups (unfavourable vs intermediate vs favourable, 7% 95%-CI 4–10% vs 7% 95%-CI 5–9% vs 6% 95%-CI 3–9%; hazard ratio, 0.94; 95%-CI, 0.50–1.78; P = 0.847; Fig. 2b).
Similar results were obtained when donor ct-KIR score groups were compared with regard to incidence of non-relapse mortality (NRM). In univariable analysis NRM incidence proved lowest in the unfavourable donor group (Fig. 2c). However, this difference between the donor ct-KIR score groups was not statistically significant, an observation that was confirmed in multivariable analysis of NRM (hazard ratio, 1.41; 95%-CI, 0.81–2.45; P = 0.222; Table 3).
Impact of donor Ct-KIR score on event-free survival
Finally, we investigated the effect of donor ct-KIR status on event-free survival (EFS). In univariable analysis we observed a beneficial effect of donor ct-KIR status with 5-year EFS of 63% 95%-CI 57–69% vs. 68% 95%-CI 64–72% vs. 75% 95%-CI 70–80% for the unfavourable vs. intermediated vs. favourable ct-KIR score donor group (hazard ratio, 0.76; 95%-CI, 0.57–1.01; P = 0.055; Fig. 2d). This was verified in multivariable analysis where donor ct-KIR status proved associated with improved EFS (hazard ratio for one-point increment in ct-KIR score, 0.76; 95%-CI, 0.57–1.01; P = 0.059; Table 3). Although no other donor factor had an effect on EFS, disease recurrence risk was affirmed as an adverse disease characteristic (hazard ratio, 1.91; 95%-CI, 1.26–2.89; P = 0.002).
The importance of mismatching between KIR on donor NK cells and their HLA-encoded ligands in the recipient for control of myeloid leukaemia, also referred to as NK cell alloreactivity, was initially described for haploidentical HSCT [23, 24]. In the more prevailing setting of T cell-replete transplantation from HLA-matched donors, however, numerous studies yielded conflicting results on the impact of NK cell alloreactivity mainly attributed to widely varying degree of T cell depletion . Notably, NK cell alloreactivity, defined by KIR – KIR ligand mismatch [26, 27], had no impact on relapse incidence in our analysis (Fig. S6). More recent analyses in adult AML suggested a critical role for conserved KIR haplotype structures consisting of centromeric and telomeric KIR linkage groups, the inhibitory KIR2DL1 and the stimulatory KIR2DS1 in the donor [4,5,6, 28, 29]. In lymphoid malignancy, similar reports in HLA-matched HSCT yielded rather inconclusive results in adults [5, 15, 28,29,30,31] or in the case of paediatric ALL, are entirely missing. The only existing data in the paediatric setting comes from a dual centre study of 85 paediatric ALL patients undergoing haploidentical transplantation showing a beneficial impact of KIR B haplotypes as defined by Cooley  on relapse incidence . As ALL represents the most prominent indication for paediatric allogeneic HSCT the present study focussed on the role of donor KIR haplotype polymorphism in this scenario in a candidate approach. Aforementioned studies from Cooley et al. [5, 6, 32] were conducted in adult patients whereas our patient cohort consisted of a pure paediatric cohort. While ALL occurs in adults and children, there is evidence for different susceptibilities of childhood ALL and adult ALL towards NK cell-mediated leukaemia control that have been attributed to specific HLA class I downregulation in paediatric ALL . These differences in the role of KIR B genes between adults and children, and their underlying disease indicate that investigations in other diseases are highly warranted.
The setup of the ALL-SCT-BFM-2003 trial  enabled retrospective analysis of KIR polymorphism in a cohort of 317 children, homogeneously treated with regard to transplant indication, donor selection, pre-transplant conditioning and GvHD prophylaxis within a prospective, multicentric clinical trial. In full concordance with previous observations in adult AML [4,5,6, 30, 34, 35] presence of centromeric group B motifs was significantly associated with relapse protection and was thus validated in our paediatric patient cohort. We extend these findings by showing that the presence of telomeric group B motifs is detrimental for leukaemia control. The ct-KIR score, an algorithm that accommodates this dichotomous role of centromeric versus telomeric KIR B motifs enabled the distinction between three different groups including a favourable donor group that exhibited a substantially decreased relapse rate and improved event-free survival. This protective effect was not associated with any detrimental impact on other transplant outcomes including acute and chronic GvHD as well as non-relapse mortality. In multivariable analyses adjusting for disease, transplant and donor characteristics disease recurrence risk, reflected by remission status at transplant and time to first relapse, and donor ct-KIR score proved to represent the only prognostic factors for relapse incidence and event-free survival. Moreover, one thousand randomly selected bootstrap resamples of the >300 individuals in our analysis were utilised to re-assess the cumulative incidence of relapse for the subgroups of patients identified by donor ct-KIR score and validated the robust predictive value of the donor ct-KIR score on disease control (Fig. 1e).
As our analysis was retrospective in nature it may be subject to inherent biases. The retrospective design at the same time ascertains that transplant physicians were unaware of the donor KIR status at donor selection. Moreover, ct-KIR status, segregating independently from other donor characteristics including HLA genotype, did not correlate with patient, disease, and transplant characteristics, hence approving its random distribution in our study cohort.
Based on the paucity of paediatric studies donor selection criteria and their hierarchy in HSCT for childhood malignancy largely rely on registry-based analyses comprising a wide variety of underlying diseases, transplant modalities and patient ages [1, 18,19,20, 36, 37]. In multivariable analyses incorporating those characteristics that are currently employed for donor selection donor ct-KIR score proved the only donor factor associated with improved leukaemia control and event-free survival (Table 3). The question whether a patient would benefit from a favourable 9/10 MUD compared to an unfavourable 10/10 donor is of high clinical relevance. However, in order to be able to analyse more donor characteristics in a multivariable model, a higher number of patients would be needed.
The detrimental role of telomeric KIR B genes is difficult to interpret as far more genes are involved in telomeric haplotypes compared to centromeric haplotypes [38, 39]. Moreover, in contrast to the centromeric part, various ligands of receptors encoded by telomeric KIR genes are completely unknown . Thus, any biologic rationale behind the telomeric KIR B effects would remain speculative. Clearly, more research is warranted in this area.
Recently, MRD level prior to HSCT has emerged as an important prognostic factor for relapse risk . As MRD level prior to HSCT was not routinely investigated within the ALL-SCT-BFM-2003 trial, data were not available for the entire study population. To nevertheless account for the potential prognostic impact of MRD on our observations we evaluated the effect of ct-KIR score within 171 patients with available MRD data. Both univariate and multivariate analyses confirmed an MRD-independent prognostic impact of donor ct-KIR score on relapse incidence (Table S2/ Fig. S7).
In conclusion, within the attendant limitations of a retrospective analysis lacking formal validation in an independent patient cohort our data indicate that in children with ALL undergoing HLA-matched allogeneic stem-cell transplantation after myeloablative conditioning donor selection with a particular focus on the ct-KIR score holds promise to prospectively reduce relapse rates without detrimental effects on GvHD and non-relapse mortality. This may prove of particular clinical relevance if more than one appropriate donor with regard to conventional donor selection criteria is identified. Given the major hurdles to perform a randomised trial on KIR status-based donor selection in the paediatric setting studies such as ours are one means of providing informed guidance regarding donor selection. As post-transplant leukaemia relapses represent the major obstacle to treatment success and are associated with dismal prognosis our results may stimulate critical review of currently applied donor selection criteria for this group of patients. With more than 25 million registered stem-cell donors worldwide and the growing proportion of instantly available high resolution HLA- and KIR genotyping results in many registries the identification of an optimal donor based on donor ct-KIR score – representing 26% of all donors in our study and more than 31% of the Caucasoid population  seems a clinically feasible strategy which could readily be implemented into clinical care without any discernable risk.
Peters C, Schrappe M, von Stackelberg A, Schrauder A, Bader P, Ebell W, et al. Stem-cell transplantation in children with acute lymphoblastic leukemia: A prospective international multicenter trial comparing sibling donors with matched unrelated donors-The ALL-SCT-BFM-2003 trial. J Clin Oncol. 2015;33:1265–74.
Babor F, Fischer JC, Uhrberg M. The role of KIR genes and ligands in leukemia surveillance. Front Immunol. 2013;4:27.
Manser AR, Weinhold S, Uhrberg M. Human KIR repertoires: shaped by genetic diversity and evolution. Immunol Rev. 2015;267:178–96.
Cooley S, Trachtenberg E, Bergemann TL, Saeteurn K, Klein J, Le CT, et al. Donors with group B KIR haplotypes improve relapse-free survival after unrelated hematopoietic cell transplantation for acute myelogenous leukemia. Blood. 2009;113:726–32.
Cooley S, Weisdorf DJ, Guethlein LA, Klein JP, Wang T, Le CT, et al. Donor selection for natural killer cell receptor genes leads to superior survival after unrelated transplantation for acute myelogenous leukemia. Blood. 2010;116:2411–9.
Cooley S, Weisdorf DJ, Guethlein LA, Klein JP, Wang T, Marsh SG, et al. Donor killer cell Ig-like receptor B haplotypes, recipient HLA-C1, and HLA-C mismatch enhance the clinical benefit of unrelated transplantation for acute myelogenous leukemia. J Immunol. 2014;192:4592–4600.
Oevermann L, Michaelis SU, Mezger M, Lang P, Toporski J, Bertaina A, et al. KIR B haplotype donors confer a reduced risk for relapse after haploidentical transplantation in children with ALL. Blood. 2014;124:2744–7.
Sobecks RM, Wang T, Askar M, Gallagher MM, Haagenson M, Spellman S, et al. Impact of KIR and HLA genotypes on outcomes after reduced-intensity conditioning hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2015;21:1589–96.
Locatelli F, Merli P, Pagliara D, Li Pira G, Falco M, Pende D et al. Outcome of children with acute leukemia given HLA-haploidentical HSCT after alphabeta T-cell and B-cell depletion. Blood https://doi.org/10.1182/blood-2017-04-779769 2017.
Uhrberg M, Parham P, Wernet P. Definition of gene content for nine common group B haplotypes of the Caucasoid population: KIR haplotypes contain between seven and eleven KIR genes. Immunogenetics. 2002;54:221–9.
Vilches C, Castano J, Gomez-Lozano N, Estefania E. Facilitation of KIR genotyping by a PCR-SSP method that amplifies short DNA fragments. Tissue Antigens. 2007;70:415–22.
Prentice RL, Kalbfleisch JD, Peterson AV Jr., Flournoy N, Farewell VT, et al. The analysis of failure times in the presence of competing risks. Biometrics. 1978;34:541–54.
Gray R. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat. 1988;16:1141–54.
Fine J, Gray R. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509.
Bachanova V, Weisdorf DJ, Wang T, Marsh SG, Trachtenberg E, Haagenson MD, et al. Donor KIR B genotype improves progression-free survival of non-Hodgkin lymphoma patients receiving unrelated donor transplantation. Biol Blood Marrow Transplant. 2016;22:1602–7.
Lee SJ, Klein J, Haagenson M, Baxter-Lowe LA, Confer DL, Eapen M, et al. High-resolution donor-recipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood. 2007;110:4576–83.
Woolfrey A, Klein JP, Haagenson M, Spellman S, Petersdorf E, Oudshoorn M, et al. HLA-C antigen mismatch is associated with worse outcome in unrelated donor peripheral blood stem cell transplantation. Biol Blood Marrow Transplant. 2011;17:885–92.
Pidala J, Lee SJ, Ahn KW, Spellman S, Wang HL, Aljurf M, et al. Nonpermissive HLA-DPB1 mismatch increases mortality after myeloablative unrelated allogeneic hematopoietic cell transplantation. Blood. 2014;124:2596–606.
Kollman C, Spellman SR, Zhang MJ, Hassebroek A, Anasetti C, Antin JH, et al. The effect of donor characteristics on survival after unrelated donor transplantation for hematologic malignancy. Blood. 2016;127:260–7.
Ljungman P, Brand R, Hoek J, de la Camara R, Cordonnier C, Einsele H, et al. Donor cytomegalovirus status influences the outcome of allogeneic stem cell transplant: a study by the European group for blood and marrow transplantation. Clin Infect Dis. 2014;59:473–81.
Bader P, Kreyenberg H, von Stackelberg A, Eckert C, Salzmann-Manrique E, Meisel R, et al. Monitoring of minimal residual disease after allogeneic stem-cell transplantation in relapsed childhood acute lymphoblastic leukemia allows for the identification of impending relapse: results of the ALL-BFM-SCT 2003 trial. J Clin Oncol. 2015;33:1275–84.
Lee SJ, Klein JP, Barrett AJ, Ringden O, Antin JH, Cahn JY, et al. Severity of chronic graft-versus-host disease: association with treatment-related mortality and relapse. Blood. 2002;100:406–14.
Ruggeri L, Capanni M, Urbani E, Perruccio K, Shlomchik WD, Tosti A, et al. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science. 2002;295:2097–2100.
Ruggeri L, Mancusi A, Capanni M, Urbani E, Carotti A, Aloisi T, et al. Donor natural killer cell allorecognition of missing self in haploidentical hematopoietic transplantation for acute myeloid leukemia: challenging its predictive value. Blood. 2007;110:433–40.
Mehta RS, Rezvani K. Can we make a better match or mismatch with KIR genotyping? Hematology Am Soc Hematol Educ Program. 2016;2016:106–18.
Leung W. Use of NK cell activity in cure by transplant. Br J Haematol. 2011;155:14–29.
Leung W, Iyengar R, Turner V, Lang P, Bader P, Conn P, et al. Determinants of antileukemia effects of allogeneic NK cells. J Immunol. 2004;172:644–50.
Venstrom JM, Pittari G, Gooley TA, Chewning JH, Spellman S, Haagenson M, et al. HLA-C-dependent prevention of leukemia relapse by donor activating KIR2DS1. New Engl J Med. 2012;367:805–16.
Bari R, Rujkijyanont P, Sullivan E, Kang G, Turner V, Gan K, et al. Effect of donor KIR2DL1 allelic polymorphism on the outcome of pediatric allogeneic hematopoietic stem-cell transplantation. J Clin Oncol. 2013;31:3782–90.
Stringaris K, Adams S, Uribe M, Eniafe R, Wu CO, Savani BN, et al. Donor KIR Genes 2DL5A, 2DS1 and 3DS1 are associated with a reduced rate of leukemia relapse after HLA-identical sibling stem cell transplantation for acute myeloid leukemia but not other hematologic malignancies. Biol Blood Marrow Transplant. 2010;16:1257–64.
Kroger N, Binder T, Zabelina T, Wolschke C, Schieder H, Renges H, 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;82:1024–30.
Cooley S, Miller JS. “Self”-reflection by KIR. Blood. 2009;114:2–3.
Pfeiffer M, Schumm M, Feuchtinger T, Dietz K, Handgretinger R, Lang P. Intensity of HLA class I expression and KIR-mismatch determine NK-cell mediated lysis of leukaemic blasts from children with acute lymphatic leukaemia. Br J Haematol. 2007;138:97–100.
Chen C, Busson M, Rocha V, Appert ML, Lepage V, Dulphy N, et al. Activating KIR genes are associated with CMV reactivation and survival after non-T-cell depleted HLA-identical sibling bone marrow transplantation for malignant disorders. Bone Marrow Transplant. 2006;38:437–44.
Symons HJ, Leffell MS, Rossiter ND, Zahurak M, Jones RJ, Fuchs EJ. Improved survival with inhibitory killer immunoglobulin receptor (KIR) gene mismatches and KIR haplotype B donors after nonmyeloablative, HLA-haploidentical bone marrow transplantation. Biol Blood Marrow Transplant. 2010;16:533–42.
Furst D, Muller C, Vucinic V, Bunjes D, Herr W, Gramatzki M, et al. High-resolution HLA matching in hematopoietic stem cell transplantation: a retrospective collaborative analysis. Blood. 2013;122:3220–9.
Horan J, Wang T, Haagenson M, Spellman SR, Dehn J, Eapen M, et al. Evaluation of HLA matching in unrelated hematopoietic stem cell transplantation for nonmalignant disorders. Blood. 2012;120:2918–24.
Wilson MJ, Torkar M, Haude A, Milne S, Jones T, Sheer D, et al. Plasticity in the organization and sequences of human KIR/ILT gene families. Proc Natl Acad Sci USA. 2000;97:4778–83.
Martin AM, Kulski JK, Gaudieri S, Witt CS, Freitas EM, Trowsdale J, et al. Comparative genomic analysis, diversity and evolution of two KIR haplotypes A and B. Gene. 2004;335:121–31.
van Bergen J, Trowsdale J. Ligand specificity of Killer cell Immunoglobulin-like receptors: a brief history of KIR. Front Immunol. 2012;3:394.
We thank all parents who gave their consent to use the biological material from their minors. This work was supported by the Deutsche Krebshilfe e.V. (M.U., R.M., and A.B.) (project 110351), the Forschungskommission of the Medical Faculty of the Heinrich-Heine University Düsseldorf (F.B.), and the TRANSAID - Stiftung fuer krebskranke Kinder (F.B.). The current affiliation of M.Si. is Department of Pediatrics and Mother and Child Center, Hospital Neuwerk, Mönchengladbach, Germany.
F.B., A.M., J.M., C.E. and N.S. performed the experiments; M.S., M.A., G.C., T.F., B.G., T.G., P.A.H., B.K., P.L., M.M., I.M., J.M., L.O., H.P., F.R.S., M.S., D.S., B.S., W.W., G.E., M.Z., M.S., A.B. and P.B. provided samples and conducted data the analysis and interpretation, and participated in patient care; E.G. and U.P. designed and performed the bioinformatic analyses; F.B., C.P., M.U. and R.M. wrote the manuscript, designed and directed the study; and all authors contributed to the research and approved the final manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Babor, F., Peters, C., Manser, A.R. et al. Presence of centromeric but absence of telomeric group B KIR haplotypes in stem cell donors improve leukaemia control after HSCT for childhood ALL. Bone Marrow Transplant 54, 1847–1858 (2019). https://doi.org/10.1038/s41409-019-0543-z
Decreased iKIR-HLA C Pair Confers Worse Clinical Outcomes for Patients With Myeloid Disease Receiving Antithymocyte Globulin-Based Haploidentical Hematopoietic Stem Cell Transplantation
Frontiers in Immunology (2021)
Investigation of donor KIR content and matching in children undergoing hematopoietic cell transplantation for acute leukemia
Blood Advances (2020)
Influence of KIR and NK Cell Reconstitution in the Outcomes of Hematopoietic Stem Cell Transplantation
Frontiers in Immunology (2020)
Phenotypic and Functional Characterization of NK Cells in αβT-Cell and B-Cell Depleted Haplo-HSCT to Cure Pediatric Patients with Acute Leukemia
The impact of donor full‐length KIR2DS4 in the development of acute and chronic GVHD after unrelated allogeneic HSCT
Pediatric Transplantation (2020)