Selecting Double Cord Blood Units for Adults with Hematological Malignancies: Impact of ABO, HLA and Cell Dose on Outcomes after Double Cord Blood Transplants a CBC-Cellular Therapy & Immunobiology Working Party, EBMT and Eurocord Study

Despite the increased inventory of "rich" and more diverse HLA cord blood units (CBUs), most adults are still transplanted with double unrelated CBUs. Many studies have described criteria for CBU selection for single UCBT (SUCBT) based on cell dose, low and high resolution (LR and HR) HLA typing, KIR and NIMA. However, the criteria based on these factors for selecting double CBUs are still not as well defined. Moreover, most of SUCBT are performed in children whereas double UCBT (DUCBT) are commonly used for adults, therefore CBUs selection criteria may be different between in these two settings.

With the aim of studying the impact of CBUs, patient and transplantation-related factors in DUCBT, we conducted a retrospective registry-based study in a large series of adults transplanted with DUCBT in EBMT centers, for whom we had available data on cell dose and HLA for both CBUs. Overall, 934 adults met the eligibility criteria: 55% had acute leukemia, 21% lymphoid mature malignancies, 10% MDS and 14% MPD or multiple myeloma. Patients' median age at DUCBT was 47 years (y) and median weight was 71 kg. Median time from diagnosis to DUCBT was 17 months. Conditioning regimen was myeloablative (MAC) in 32% and antithymocyte globulin (ATG) was used in 24% of the cases. GvHD prophylaxis consisted mostly of cyclosporine+mycophenolate mofetil. Median number of TNC at collection and infusion were 4.9 x10⁷/kg (n=934) and 3.6x10⁷/kg (n=735), respectively. Median number of collected and infused CD34 cells were 1.9x10⁵/kg (n=893) and 1.6 x10⁵/kg (n=695), respectively. Number of HLA disparities was defined according to the CBU with the highest degree of mismatches with the recipient, based on LR typing for HLA-A and -B and HR typing for -DRB1. Overall, 1% of the patients received a 6/6, 26% a 5/6, 67% a 4/6 and 6% a 3/6 HLA-matched DUCBT. ABO match was also classified according to the CBU with the highest incompatibility with the recipient: 18% of the patients received two ABO compatible CBUs, 30% at least one CBU with minor ABO incompatibility and 52% received at least one CBU with a major ABO incompatibility.

At day 60, estimated absolute neutrophil count (ANC>500) recovery was 87%. In multivariate analysis (MVA) greater ANC recovery was associated with TNC dose at collection≥ 4.9x10⁷/kg (HR:0.66; p=0.002), as well as female recipient, negative CMV serology and other diagnosis than MDS. Impact of CD34 cell dose was observed in univariate analysis: probability of ANC recovery was 91% if ≥1.9x10⁵/kg versus 84% for a lower dose (p<0.001). At day-100, the incidence of acute GvHD (aGvHD) grade II-IV was 42% and III-IV was 20%. In MVA, increased aGVHD (II-IV) was associated with absence of ATG, use of MAC and HLA-match ≥4/6 (HR: 1.51, p=0.007), whereas increased aGVHD (III-IV) was associated with recipient's age (>47y), advanced disease, HLA-match ≥4/6 (HR: 1.55; p=0.048) and ABO major incompatibility (HR:1.44; p=0.039). The 2y NRM was 35%. Factors associated with higher NRM were positive CMV serology, advanced disease, use of ATG, ABO incompatibility (HR:1.55; p=0.023) and HLA-match ≥4/6 (HR:1.39; p=0.046). Relapse incidence (RI) at 2y was 20% and none of the CBU related factors impacted on RI. Overall Survival (OS) at 2y was 47%. In MVA, positive CMV serology, diagnosis of MDS and use of ATG were associated with decreased OS. Number of HLA disparities and degree of ABO compatibility were also associated with survival: 2y OS was 45% for patients transplanted with at least one ≥4/6 HLA-matched CBU, and 54% for those transplanted with both 6/6 or at least one 5/6 HLA-matched CBU (HR:1.26; p=0.05). Interestingly, OS at 2y was 57% for patients transplanted with both ABO compatible CBUs, 47% for those transplanted with at least one CBU with ABO minor incompatibility and 39% for those transplanted with at least one major ABO incompatibility (HR:1.16; p=0.029). In MVA for DFS, positive CMV serology, advanced disease, use of ATG and ABO incompatibility (HR:1.46;p=0.004) were associated with decreased DFS.

This analysis suggests that a dose at collection of TNC ≥ 4.9x10⁷/kg and/or of CD34 ≥ 1.9x10⁵/kg, as well as better ABO and HLA compatibility are important factors that should be included in the selection algorithm for CBUs in DUCBT.