Search

Donor KIR and recipient HLA combinations that minimize inhibition and favor activation of the NK repertoire are associated with improved outcomes after allogeneic hematopoietic cell transplantation (HCT) in patients with myeloid neoplasia. We prospectively evaluated a weighted donor ranking algorithm designed to prioritize HLA-compatible unrelated donors (URDs) with weak inhibitory KIR3DL1/HLA-Bw4 interaction, followed by donors with nontolerized activating KIR2DS1, and finally those with KIR centromeric B haplotype. During donor evaluation, we performed KIR genotyping and ranked 2079 URDs for 527 subjects with myelodysplastic syndrome (MDS) or acute myelogenous leukemia (AML). Among all patients, 394 (75%) had at least 1 KIR-advantageous donor, and 263 (50%) underwent HCT. In patients with AML, KIR3DL1 weak inhibition provided protection from relapse. Compared with KIR3DL1-Weak Inhibiting donors, KIR3DL1-Noninteracting donors were associated with increased risk of relapse (HR, 2.97; 95% CI, 1.33-6.64; P = .008) and inferior event-free survival (EFS; HR, 2.14; 95% CI, 1.16-3.95; P = .015). KIR3DL1-Strong Inhibiting donors were associated with HR, 1.65 (95% CI, 0.66-4.08; P = .25) for AML relapse and HR, 1.6 (95% CI, 0.81-3.17; P = .1) for EFS when compared with the use of KIR3DL1-weak inhibiting donors. Donor KIR2DS1/HLA-C1 status and centromeric KIR haplotype-B content were not associated with decreased risk of AML relapse. There was no benefit to KIR-based donor selection in patients with MDS. This study demonstrates that donor KIR typing is feasible, and prioritization of donors with certain KIR3DL1 genotypes may confer a protection from relapse after HCT in patients with AML.

Introduction Optimal CB unit selection incorporates TNC & CD34+ cell dose & 8-allele donor-recipient HLA-match. How graft characteristics for these parameters vary by patient (pt) race/ ethnicity, however, is unknown. Methods We analyzed cryopreserved TNC & CD34+ cell content, the cell dose incorporating pt weight, & HLA-match of infused & backup (reserved but not shipped) units by pt ancestry in CB transplant (CBT) recipients transplanted 7/2013-6/2018. Units were chosen based on banking practices, TNC & CD34+ dose & 4-6/6 & 8-allele HLA-match. Patient (pt) racial/ ethnic origins were prospectively obtained & grouped per Barker J. et al, BBMT 2010. Results The characteristics of 544 units chosen for 144 diverse CBT recipients by pt ancestry are shown (Table 1). 286 units were infused (142 doubles, 2 singles) & pts had a total of 258 backup units (0-2/pt). The median TNC content of units for Europeans was higher than for Non-Europeans (216 vs 197, p = 0.002). Units for NW Europeans had the highest median TNC content with lowest in units for Africans & White Hispanics. Units chosen for Europeans also had a higher median CD34+ cell content than Non-Europeans (160 vs 132, p = 0.007). Units for NW Europeans had the highest & units for Africans the lowest median CD34+ content. Median TNC/kg was similar in units for European & Non-European pts (2.7 vs 2.5, p = 0.11) & differences between groups were not marked. However, there was a trend for units for Europeans having a higher median CD34+ dose than units for Non-Europeans (1.9 vs 1.7, p = 0.057) with disparity between groups. Although the median 8 allele HLA-match was 5/8 for units chosen for both European & Non-European pts, units identified for non-Europeans were more mismatched (p = 0.017). When the 286 transplanted units (considered the best) were analyzed, the median TNC & CD34+ contents of units were higher for Europeans than non-Europeans (232 vs 216, p = 0.022, & 181 vs 152, p = 0.009). As shown in Table 2, while there were no overall differences in the median TNC & CD34+ doses of units for Europeans & non-Europeans, differences in cell content combined with differences in weights resulted in TNC & CD34+ dose disparities within ancestry groups. For example, NW Europeans (high weight, high CD34+ content) had the highest CD34+ doses; lower content in Asian pts was partially compensated for by low weight; African pts (high weight, low content) had the lowest doses. Finally, there was marked differences in HLA-match distribution by ancestry group: eg only 23% of units for NW Europeans were 3-4/8 vs 40% for S. Europeans, 46% for white Hispanics & 51% for Africans. Conclusions CB extends transplant access to minorities but differences in cellular content translates to many ancestry groups pts having access to only lower dosed units. There are also marked racial/ ethnic disparities in HLA-match. This data supports continued funding of banks to further increase the CB inventory.

Introduction: Optimal CB unit selection guidelines recommend consideration of CD34+ cell dose & 8-allele donor-recipient HLA-match. How graft characteristics for these parameters vary by patient (pt) race/ ethnicity, however, is not known. Methods: We analyzed the infused graft & back-up unit cryopreserved total nucleated cell (TNC) x 107 & CD34+ x 105 cell content, the cell dose (incorporating pt weight), & 4-6/6 & 8-allele HLA-match by pt ancestry in CB transplant (CBT) recipients transplanted 1/2014-6/2018. Units were chosen based on banking practices (e.g. RBC depleted, standard cryo volumes), TNC & CD34+ dose & 4-6/6 & 8-allele HLA-match with dose usually taking priority over match given pt size at our center. The analysis included transplanted units (considered the best choice) & the next best high resolution typed back-up units (reserved but not shipped). Pt racial/ ethnic origins were prospectively obtained by detailed family history & grouped as previously described (Barker J. et al. BBMT 2010). Results: The characteristics of 513 units chosen for 136 CBT recipients by pt ancestry are shown (Table 1). Pts had highly diverse origins including 70 (51%) non-Europeans. The 513 units included 270 units infused as the graft (134 doubles & 2 singles) & 243 back-up units (109 pts had 2 back-ups, 25 pts had one & 2 had none). Thus, 4 best units were analyzed in 109 pts (all double unit recipients), 3 best in 25 pts (all doubles), & 1 unit in 2 pts (both singles). The median weight of the 136 pts was 81 kg. Asian pts (median 68 kg) had a lower weight than other groups. The median TNC content of units for the 66 European pts was higher than that for the 70 Non-Europeans (218 vs 196, p = 0.004). Units chosen for Northwestern (NW) Europeans had the highest median TNC content (235) with lower TNC content in units for Southern Europeans (202), Asian (193), African (191) & White Hispanic (189) pts. Units chosen for European pts also had a higher median CD34+ cell content (162) than Non-Europeans (138), p = 0.004. NW Europeans had units with a higher median CD34+ content (198) & the lowest CD34+ content were those for African (124) & Middle Eastern pts (124). When patient weight was considered, median TNC/kg dose per unit was similar in European and Non-European pts (2.7 vs 2.6, p = NS). Units for NW Europeans had the highest median TNC dose (3.0) whereas those for African pts had the lowest TNC dose (2.4). Units for Europeans had a higher median CD34+ dose (2.0) than Non-Europeans (1.7) although this difference was not significant (p = 0.15). Additionally, similar to TNC dose, median CD34+ dose was highest in units for NW European pts (2.2) & lowest in units chosen for African pts (1.5). 89% of chosen units were 4/6 HLA-matched with no differences between Europeans & non-Europeans. Furthermore, the median 8 allele HLA-match was 5/8 (range 2-8/8) with no overall differences between units for Europeans and Non-Europeans (p = NS). When only transplanted units were analyzed (Table 2), the median TNC & CD34+ contents were significantly lower in non-Europeans than Europeans (238 vs 216, p = 0.01 & 184 vs 160, p = 0.016). Overall, however, units received by Europeans vs non-European pts had similar TNC & CD34+ doses (p = NS). However, differences in the CD34+ content combined with differences in pt weights resulted in disparities in CD34+ doses by ancestry sub-group. NW Europeans (high weight, high CD34+ content) received the best CD34+ doses; lower CD34+ content in Asian pts was compensated for by their lower weight. African pts (high weight, low CD34+ content) received the lowest CD34+ doses. The median 8 allele HLA-match for all was 5/8 (range 3-8/8) with the exception of African pts [median 4/8 (range 3-7/8)]. Moreover, while 108 (40%) of transplanted units were 3-4/8 HLA-matched overall, there were marked differences between pt sub-groups with only 23% of units for NW Europeans being 3-4/8 vs 42% for southern Europeans, 46% for white Hispanics & 53% for Africans. Conclusions: While CB significantly extends transplant access to racial & ethnic minorities, differences in cellular content translates to many minority pts receiving lower dosed units. There are also marked racial/ ethnic differences in HLA-match grade with African pts the most likely to receive highly mismatched units. This data supports ongoing funding of public CB banks to further increase the inventory of high dosed & better matched units for all but especially racial & ethnic minority pts. Disclosures Shah: Janssen: Research Funding; Amgen: Research Funding. Kernan:National Cancer Institute: Research Funding.