Impact of total body irradiation on successful neutrophil engraftment in unrelated bone marrow or cord blood transplantation

Received: 3 November 2016 Revised: 16 November 2016 Accepted: 21 November 2016 DOI 10.1002/ajh.24613 RESEARCH ARTICLE Impact of total body irradiation on successful neutrophil engraftment in unrelated bone marrow or cord blood transplantation Hideki Nakasone 1 Fuji Shigeo 2 Kimikazu Yakushijin 3 Makoto Onizuka 4 Akihito Shinohara 5 Kazuteru Ohashi 6 Koichi Miyamura 7 Naoyuki Uchida 8 Minoko Takanashi 9 Tatsuo Ichinohe 10 Yoshiko Atsuta 11,12 Takahiro Fukuda 2 Masao Ogata, 13 On Behalf of the Complication Working Group of Japanese Society for Hematopoietic Cell Transplantation 1 Division of Hematology, Saitama Medical Center, Jichi Medical University, Saitama, Japan; 2 Department of Hematopoietic Stem Cell Transplantation, National Cancer Center Hospital, Tokyo, Japan; 3 Department of Medical Oncology and Hematology, Kobe University Hospital, Kobe, Japan; 4 Department of Hematology and Oncology, Tokai University School of Medicine, Isehara, Japan; 5 Department of Hematology, Tokyo Women s Medical University, Tokyo, Japan; 6 Hematology Division, Tokyo Metropolitan Cancer and Infectious Diseases Center, Komagome Hospital, Tokyo, Japan; 7 Department of Hematology, Japanese Red Cross Nagoya First Hospital, Nagoya, Japan; 8 Department of Hematology, Toranomon Hospital, Tokyo, Japan; 9 Blood Service Headquarters, Japanese Red Cross Society, Tokyo, Japan; 10 Department of Hematology and Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan; 11 Japanese Data Center for Hematopoietic Cell Transplantation, Nagoya, Japan; 12 Department of Healthcare Administration, Nagoya University Graduate School of Medicine, Nagoya, Japan; 13 Department of Medical Oncology and Hematology, Oita University Faculty of Medicine, Oita, Japan Correspondence Hideki Nakasone MD/PhD, Division of Hematology, Saitama Medical Center, Jichi Medical University, 1-847 Amanuma-cho Omiya-ku, Saitama, 330-8503, Japan. Email: nakasone-tky@umin.ac.jp Funding Information This work was supported in part by a Research Grant for Allergic Disease and Immunology from the Japanese Ministry of Health, Labor, and Welfare (Y.A.). H.N. was a recipient of research awards from Jichi Medical University. Abstract Total body irradiation (TBI) has been thought to promote donor cell engraftment in allogeneic hematopoietic cell transplantation (HCT) from alternative donors. However, recent progress in HCT strategies may affect the clinical significance of TBI on neutrophil engraftment. With the use of a Japanese transplant registry database, we analyzed 3933 adult recipients (>15 y.o.) who underwent HCT between 2006 and 2013 from an 8/8 HLA-matched unrelated bone marrow donor (MUD, n 5 1367), an HLA-mismatched unrelated bone marrow donor (MMUD, n 5 1102), or unrelated cord blood (CBT, n 5 1464). Conditioning regimens were divided into five groups: High-TBI-(>8Gy), Low-TBI- (8Gy), and no-tbi-myeloablative conditioning (MAC), and Low-TBIand no-tbi-reduced-intensity conditioning (RIC). In both MUD and MMUD, neutrophil engraftment rate was >90% in each of the five conditioning groups, and TBI was not associated with prompt neutrophil engraftment in multivariate analyses. Conversely, in CBT, TBI regimens had a higher rate of day-30 neutrophil engraftment than no-tbi-regimens: 78% in High-TBI-MAC, 83% in Low-TBI-MAC, and 76% in Low-TBI-RIC versus 65% in No-TBI-MAC, and 68% in No-TBI-RIC (P <.001). Multivariate analyses in CBT demonstrated that TBI-regimens were significantly associated with a higher rate of neutrophil engraftment. Subsequently focusing on CBT patients alone, TBI-regimens were significantly associated with a higher rate of neutrophil engraftment in patients who received CBT with a 4/6 or less HLA allele-match, or who had anti-hla antibodies. In summary, TBI-regimens had no impact on neutrophil engraftment in the current practice of unrelated bone marrow transplantation. However, in CBT, TBI is still necessary to enhance engraftment. Conflicts of Interest: The authors report no potential competing conflicts of interest. Am J Hematol 2017; 92: 171 178 wileyonlinelibrary.com/journal/ajh VC 2016 Wiley Periodicals, Inc. 171

172 1 INTRODUCTION Allogeneic hematopoietic cell transplantation (HCT) has been developed as a curative treatment for hematological malignancies. HCT recipients will frequently experience various adverse events, and all of them initially have to achieve donor cell engraftment during the first month following HCT to acquire an allogeneic anti-leukemia/lymphoma effect. Historically, allogeneic HCT has spread widely since the initial success of high-dose total body irradiation (TBI) with cyclophosphamide (CY) 50-year ago. Since then, TBI with CY has been the standard myeloablative conditioning (MAC) regimen. 1 Conversely, a higher dose of irradiation has been shown to result in increased organ toxicities, 2 5 although it was also linked to a reduced risk of leukemia relapse. 1,6 Therefore, to reduce the adverse toxicities due to TBI, clinical investigators have explored alternative strategies including no-tbi or reduced-intensity conditioning (RIC). However, these strategies might raise a concern of graft failure in an unrelated transplant or HLAmismatched transplant setting due to insufficient immunosuppression, as reported in early studies. 1,7 10 Therefore, TBI has been widely incorporated into conditioning regimens because of its theoretical potential to suppress recipient-side immunity and prevent graft rejection, in addition to enhanced anti-leukemic effects. However, recent progress with conditioning strategies, donor source selection algorithms, graftversus-host disease (GVHD) prophylaxis, and various approaches to supportive care may improve sustained engraftment even in these alternative HCT settings. To the best of our knowledge, there have been no reports in large cohorts that comprehensively assessed the impact of TBI on neutrophil engraftment in the setting of unrelated HCT as a primary endpoint. Thus, the impact of TBI on neutrophil engraftment following HCT from an unrelated bone marrow (BMT) or cord blood donor was retrospectively analyzed using a Japanese transplant registry database. First, the impact of TBI on cumulative neutrophil engraftment was analyzed in each donor cohort. Next, we evaluated the effects of TBI in subgroups of the unrelated cord blood transplantation (CBT) stratified according to conditioning intensity, number of HLA-matches, and the presence of anti-hla antibodies. 2 PATIENTS AND METHODS 2.1 Definitions of categories Neutrophil engraftment was defined as the first of three sequential days when the neutrophil count was >500 3 10 6 /L. Conditioning regimens that included TBI > 8 Gy, melphalan (Mel) > 140 mg/m 2, or oral busulfan (Bu) 9 mg/kg (i.v. Bu 72 mg/kg) were classified as MAC. Other regimens were classified as RIC. 11 13 Conditioning regimens were then divided into five groups: High-TBI (> 8Gy)-MAC, Low-TBI ( 8Gy)-MAC, no-tbi-mac, Low-TBI-RIC, and no-tbi-ric. HLAmatch was defined as an 8/8 allelic match of HLA-A, -B, -C, and -DRB1 in unrelated BMT, or a 6/6 allelic match of HLA-A, -B, and -DRB1 in CBT. 2.2 Patient selection We retrospectively analyzed 3933 adult and adolescent recipients (>15- year old) who underwent their 1st HCT from an unrelated donor between 2006 and 2013: 8/8 HLA-matched unrelated bone marrow donor (MUD, n 5 1367), HLA-mismatched unrelated bone marrow donor (MMUD, n 5 1102), and unrelated cord blood (CBT, n 5 1464). HCT with unrelated peripheral blood stem cells was excluded, since it was rarely available in Japan during the evaluation period. Additionally, only standard-risk leukemia patients were included: 1st or 2nd complete remission of acute myelogenous leukemia (AML) or acute lymphoblastic leukemia (ALL), 1st or 2nd chronic phase of chronic myeloid leukemia, and myelodysplastic syndrome (MDS) other than refractory anemia with excess blasts. The data were reported to the Japan Society for Hematopoietic Cell Transplantation through the Transplant Registry Unified Management Program in 2014. 14 For recipients to be considered eligible, information on age, sex, HLA, conditioning regimen, neutrophil engraftment, and survival status at the end of follow-up were required. High-TBI-MAC included TBI and CY-based regimens, while regimens that contained TBI with cytarabine alone or etoposide alone were excluded. Low-TBI-MAC included fludarabine(flu)-based Bu or Mel regimens with TBI. No-TBI-MAC included Bu with CY, and Flu-based Bu or Mel. RIC included only Flu-based regimens. Patients who received GVHD prophylaxis other than cyclosporine or tacrolimus-based strategies were excluded. This retrospective analysis was conducted in accordance with the Declaration of Helsinki and approved by the institutional review board at Saitama Medical Center, Jichi Medical University. 2.3 Statistical analysis Categorical variables were compared using the chi-square test. The probabilities of cumulative neutrophil engraftment in each conditioning type were estimated and separately compared by Gray s methodin individual donor cohorts, where death before engraftment was considered as a competing risk. These probabilities were estimated with a 95% confidence interval (CI). The Cox proportional hazard model was used in a multivariate analysis. In the multivariate analysis, the hazard ratios (HR) of conditioning types were obtained after being adjusted for the following variables: gender, age, disease, performance status, GVHD prophylaxis, in vivo T-cell depletion, infused total nucleated cells (TNC), and use of granulocyte-colony stimulating factor (G-CSF). Statistical significance was defined as a two-tailed P-value of less than.05. All data management and statistical calculations were performed using Stata version 120 (Stata Corp., College Station, TX) and EZR, which is a graphical user interface for R (The R Foundation for Statistical Computing, version 322, Vienna, Austria) (Jichi Medical University at http://www.jichi.ac.jp/saitama-sct/saitamahp.files/statmeden.html). 15 3 RESULTS 3.1 Patient characteristics Among the 3933 patients, the donor was MUD in 1367, MMUD in 1102, and CBT in 1464 (Table 1). The patient s medianagewas47

173 TABLE 1 Patient characteristics Total (n 5 3933) (%) MUD (n 5 1367) (%) MMUD (n 5 1102) (%) CBT (n 5 1464) (%) Age 15-29 668 17% 232 17% 194 18% 242 17% 30-49 1474 37% 517 38% 459 42% 498 34% 50-1791 46% 618 45% 449 41% 724 49% Gender Female 1738 44% 597 44% 483 44% 658 45% Male 2195 56% 770 56% 619 56% 806 55% Disease AML 2132 54% 741 54% 584 53% 807 55% ALL 1231 31% 411 30% 358 32% 462 32% MDS 427 11% 171 13% 118 11% 138 9% CML 143 4% 44 3% 42 4% 57 4% Performance Status PS0-1 3757 96% 1324 97% 1059 96% 1374 94% PS2-4 161 4% 43 3% 43 4% 75 5% Missing 15 0% CMV serostatus (D/R) Pos/Pos 1177 30% 633 46% 544 49% 0 0% Pos/Neg 254 6% 141 10% 113 10% 0 0% Neg/Pos 1751 45% 378 28% 269 24% 1104 75% Neg/Neg 471 12% 118 9% 94 9% 259 18% Missing 280 7% Sex mismatch Matched 1849 47% 759 56% 616 56% 474 32% Male to Female 942 24% 399 29% 297 27% 246 17% Female to Male 689 18% 208 15% 189 17% 292 20% Missing 453 12% Conditioning HighTBI_MAC 2001 51% 673 49% 569 52% 759 52% LowTBI_MAC 222 6% 63 5% 59 5% 100 7% NoTBI_MAC 556 14% 281 21% 198 18% 77 5% LowTBI_RIC 876 22% 238 17% 188 17% 450 31% NoTBI_RIC 278 7% 112 8% 88 8% 78 5% GVHD prophylaxis CsA-based 928 24% 245 18% 127 12% 556 38% Tac-based 3005 76% 1122 82% 975 88% 908 62% T-cell depletion Without 3761 96% 1342 98% 999 91% 1420 97% With 172 4% 25 2% 103 9% 44 3% Infused TNC [310 8/ kg (310 7/ kg in CBT)] <2 706 18% 297 22% 242 22% 167 11% 2-4 2850 72% 914 67% 735 67% 1201 82% 4 369 9% 156 11% 125 11% 88 6% G-CSF Without 232 6% 113 8% 77 7% 42 3% With 3701 94% 1254 92% 1025 93% 1422 97% MUD, bone marrow transplant from an HLA-matched unrelated donor; MMUD, bone marrow transplant from an HLA-mismatched unrelated donor; CBT, unrelated cord blood transplant; AML, acute myelogenous leukemia; ALL, acute lymphoblastic leukemia; MDS, myelodysplastic syndrome; PS, performance status; CMV, cytomegalovirus; Pos, positive; Neg, negative; TBI, total body irradiation; MAC, myeloablative conditioning; RIC, reducedintensity conditioning; GVHD, graft-versus-host disease; CsA, cyclosporine; Tac, tacrolimus; TNC, total nucleated cells; G-CSF, granulocyte-colony stimulating factor. (IQR:23), 46 (IQR:21), and 49 (IQR:24) years in the MUD, MMUD, and CBT cohorts, respectively. Half of the patients received High-TBI- MAC, and TBI-regimens were used in 80% of them. Tacrolimus-based GVHD prophylaxis was frequently used in MUD- and MMUD-BMT, while cyclosporine-based prophylaxis was frequently used in CBT. Most CBT patients (n 5 1437, >98%) received single cord blood. The

174 FIGURE 1 (A) Neutrophil engraftment according to the conditioning type in the three donor cohorts. (B) Impact of TBI on neutrophil engraftment in the groups stratified according to the donor source and conditioning intensity in multivariate analyses of myeloablative and reduced-intensity conditioning cohorts. MUD, bone marrow transplant from an HLA-matched unrelated bone marrow donor; MMUD, bone marrow transplant from an HLA-mismatched unrelated bone marrow donor; CBT, unrelated cord blood donor transplant; TBI, total body irradiation; MAC, myeloablative conditioning; RIC, reduced-intensity conditioning. HR is shown, after being adjusted for gender, age, disease, performance status, GVHD prophylaxis, in vivo T-cell depletion, infused total nuclear cells, and use of granulocyte-colony stimulating factor median dose of infused TNC was 267 (IQR:127) 3 10 8 /kg in the MUD group, 264 (IQR:121) 3 10 8 /kg in the MMUD group, and 254 (IQR:085) 3 10 7 /kg in the CBT group. G-CSF was used in >90% recipients for all types of donors. The details of the patient distributions among the conditioning types are shown in Supporting Information Tables S1 and S2, according to the individual donor cohorts (MUD, MMUD, and CBT) and conditioning intensity (MAC and RIC). In the MAC group, High-TBI regimens were frequently used for younger recipients (<50 y.o.) and those with ALL in any donor cohort (Supporting Information Table S1). There was no difference in infused cell doses among the conditioning types (Novs. Low- vs. High-TBI-MAC) in any donor cohort. In the RIC group, there was no difference in age, infused cell dose or G-CSF usage between conditioning types in any donor cohort (Supporting Information Table S2). 3.2 Neutrophil engraftment in the overall MUD, MMUD, and CBT cohorts Overall neutrophil engraftment at 30 days post-hct was 96% (95%CI: 95%-97%), 95% (95%CI: 94%-96%), and 77% (95%CI: 74%-79%) in the MUD, MMUD, and CBT cohorts, respectively. In the MUD cohort, there was no difference in engraftment achievement among the types of conditioning in a univariate analysis (P 5.34, Figure 1A), and more than 95% of recipients could achieve neutrophil engraftment by 30 days post-hct in all types of conditioning [96% (95%CI: 94%-97%) in High-TBI-MAC, 97% (95%CI: 85%- 99%) in Low-TBI-MAC, 97% (95%CI: 94%-99%) in No-TBI-MAC, 97% (95%CI: 93%-98%) in Low-TBI-RIC, and 96% (95%CI:89%-98%) in No- TBI-RIC]. The median time to neutrophil engraftment were 17 days in High-TBI-MAC and Low-TBI-RIC, and 16 days in Low-TBI-MAC,

175 FIGURE 2 Cumulative neutrophil engraftment in the unrelated cord blood transplant (CBT) sub-cohorts with (A) a 6/6 or 5/6 HLA allele match and MAC (n 5 101), (B) a 4/6 or less HLA allele match and MAC (n 5 788), (C) a 6/6 or 5/6 HLA allele match and RIC (n 5 79), and (D) a 4/6 or less HLA allele match and RIC (n 5 420). (E) Multivariate analyses for the impact of TBI on neutrophil engraftment in the CBT sub-cohorts. HR is shown, after being adjusted for gender, age, disease, performance status, GVHD prophylaxis, in vivo T-cell depletion, infused total nuclear cells, and use of granulocyte-colony stimulating factor. TBI, total body irradiation; MAC, myeloablative conditioning; RIC, reduced-intensity conditioning No-TBI-MAC, and -RIC. In multivariate analyses, TBI-regimens were not associated with prompt neutrophil engraftment (Supporting Information Table S3). Rather, High-TBI-MAC was associated with delayed engraftment [HR 085 (95%CI: 072-10), P 5.048]. In the MMUD cohort, neutrophil engraftment was sufficiently achieved in all of the five conditioning types, although patients with TBI-MACs had higher probabilities of neutrophil engraftment at 30 days post-hct: 97% (95%CI: 95%-98%) in High-TBI-MAC, 97% (95%CI: 85%-99%) in Low-TBI-MAC, 95% (95%CI: 91%-97%) in No- TBI-MAC, 92% (95%CI: 87%-95%) in Low-TBI-RIC, and 93% (95%CI: 85%-97%) in No-TBI-RIC (P 5.02, Figure 1A). The median time to neutrophil engraftment were 17 days in High-TBI-MAC and Low-TBI-RIC, and 16 days in Low-TBI-MAC, No-TBI-MAC, and -RIC. However, in a multivariate analysis, there was no difference in neutrophil engraftment among the five conditioning types (Supporting Information Table S3). Conversely, in the CBT cohort, the probabilities of neutrophil engraftment at 30 days post-hct were significantly higher in TBI-regimens: 78% (95%CI: 75%-81%) in High-TBI-MAC, 83% (95%CI: 74%- 89%) in Low-TBI-MAC, and 76% (95%CI: 72%-80%) in Low-TBI-RIC versus 65% (95%CI: 53%-75%) in No-TBI-MAC, and 68% (95%CI: 56%-77%) in No-TBI-RIC (P <.001, Figure 1A). The median time to neutrophil engraftment were 23 days in High-TBI-MAC, 19 days in Low-TBI-MAC, 22 days in No-TBI-MAC, 21 days in Low-TBI-RIC, and 24 days in No-TBI-RIC. Multivariate analyses also demonstrated that TBI-regimens were significantly associated with a higher rate of neutrophil engraftment (Supporting Information Table S3). When we separately analyzed the impact of TBI in the MAC and RIC sub-cohorts, TBI had a favorable impact on neutrophil engraftment in the CBT cohort, but not in the MUD or MMUD cohort (Figure 1B). In summary, TBI did not have a favorable impact on successful neutrophil engraftment in unrelated BMT. Conversely, in the CBT cohort, TBI had a significantly favorable effect on neutrophil engraftment in both the MAC and RIC groups. Therefore, in subsequent analyses, we explored the significance of incorporating TBI into conditioning regimens, with a particular focus on the CBT sub-cohorts stratified according to HLA match, infused cell dose, and the presence of anti- HLA antibody. 3.3 Subgroup analyses according to HLA allelematch and conditioning intensity in the CBT cohort In the CBT sub-cohorts with a 6/6 or 5/6 HLA allele-match between the donor and recipient, there was no difference in neutrophil engraftment among the conditioning types in the MAC or RIC groups (Figure 2A,C). Conversely, in the CBT sub-cohort with a 4/6 or less HLA allelematch, recipients with TBI-regimens had a higher rate of neutrophil engraftment (Figure 2B,D), and TBI was significantly associated with neutrophil engraftment in both the MAC and RIC groups (Figure 2E). If we focused on the number of HLA-match of host-versus-graft (HVG) direction in the CBT group, the favorable effect of TBI on neutrophil engraftment was also observed in the subgroup with 4 or less HLA match of HVG direction, but not observed in the CBT subgroup with

176 Low-TBI: 22 (95%CI: 12-43), P 5.02]. In summary, the use of TBI might enhance neutrophil engraftment even in recipients with anti- HLA antibodies. 3.6 Impact of TBI on other outcomes in HCT from MUD, MMUD, and CBT Although the primary endpoint of this study was neutrophil engraftment, the impact of TBI on other outcomes was also assessed (Supporting Information Figure S3). In the CBT cohort, TBI-regimens tended to be associated with a lower risk of non-relapse mortality. 4 DISCUSSION FIGURE 3 Cumulative neutrophil engraftment according to TBI intensity in CBT recipients with anti-hla antibodies (n 5 173) HLA-match or one mismatch of HVG direction. In summary, TBIregimens appear to have a favorable effect on neutrophil engraftment when used along with CBT with a 4/6 or less HLA allele-match. 3.4 Subgroup analyses in CBT recipients with lower TNC or CD34-positive cell doses Next, we analyzed the impact of TBI in recipients with lower infused TNC and/or CD34-positive cell doses, which are generally considered to be risk factors for graft rejection in CBT. 16 Among recipients with a lower TNC dose (<254 3 10 7 /kg), those with TBI-regimens tended to have higher rates of neutrophil engraftment (Supporting Information Figure S1-A & S1-B). However, multivariate analyses suggested that TBI-regimens, except for Low-TBI-MAC, were not associated with prompt engraftment (Supporting Information Figure S1-C). Among recipients with lower CD34-positive cell doses (<082 3 10 5 /kg), those with TBI-regimens tended to have higher rates of neutrophil engraftment (Supporting Information Figure S2-A & S2-B). Multivariate analyses revealed that TBI-regimens were significantly associated with higher rates of neutrophil engraftment in the MAC group, but not in the RIC group (Supporting Information Figure S2-C). In summary, when CBT with low doses of infused TNC or CD34- positive cells was selected, TBI seemed to have a favorable effect on neutrophil engraftment only in specific MAC regimens. 3.5 Subgroup analyses in CBT recipients with anti-hla antibodies Furthermore, we analyzed the impact of TBI in recipients who had detectable anti-hla antibodies prior to HCT, which is also considered to be a risk factor for graft rejection. 16 18 Since only 173 recipients were reported to have anti-hla antibodies, we analyzed the impact of High-TBI, Low-TBI, and No-TBI without grouping according to the conditioning intensity (Figure 3). Multivariate analysis suggested that the use of TBI was significantly associated with increased neutrophil engraftment [HR of High-TBI: 24 (95%CI: 12-49), P 5.02; HR of This is the first study to comprehensively assess the impact of TBI on neutrophil engraftment in unrelated HCT from MUD, MMUD, and CBT using a large cohort. The results revealed that TBI-regimens were significantly associated with neutrophil engraftment in both the MAC and RIC groups of the CBT cohort. Particularly, TBI provided a benefit when recipients received HLA-mismatched CBT or had anti-hla antibodies. Additionally, in cases of CBT with a low number of CD34- positive cells, TBI may improve the probability of engraftment only in the MAC setting. Conversely, TBI did not have any favorable impact on engraftment for unrelated BMT in the current practice for standardrisk leukemia. As small studies have suggested that the risk of graft rejection might be increased in HCT from alternative donors or with the use of RIC regimens, 1,7 10 TBI has been considered to promote donor cell engraftment by suppressing recipient T cells. 1,6,19 However, the recent progress in various types of supportive care may override the concerns in HCT from alternative donors or with the use of RIC regimens. Furthermore, the clinical significance of TBI or the conditioning intensity may differ among recent HCT strategies. Thus, this study was planned to provide insight into whether the conditioning intensity or irradiation is most important for successful engraftment in HCT from alternative donors. In unrelated BMT cohorts, even if an HLA-mismatched donor was selected, neither the conditioning intensity nor TBI had a favorable impact on prompt neutrophil engraftment (Figure 1B and Supporting Information Table S3). This tendency was also observed when antigenbased HLA-mismatched HCT was analyzed instead of allele-based HLA-mismatched cohorts (data not shown). Unexpectedly, multivariate analyses showed that High-TBI-MAC was associated with delayed neutrophil engraftment in the MUD cohort (Figure 1B and Supporting Information Table S3), although the actual engraftment probabilities were fairly similar in a univariate analysis (Figure 1A). While the reason for this result is still unclear, damage to the marrow environment due to High-TBI-MAC may eventually result in impaired engraftment, as recently reported. 20 In summary, the addition of TBI as well as an increased conditioning intensity might be unnecessary to achieve neutrophil engraftment in the recent practice of unrelated BMT. Conversely, in the CBT cohort, TBI still seems to be necessary for prompt neutrophil engraftment and may be more critical than the

177 conditioning intensity (Figures 1A,B and Supporting Information Table S3). In general, graft failure in CBT is associated with HLA-disparity, 16,17,21 a lower dose of infused TNC or CD34-positive cells, 16 and the presence of anti-hla antibodies, 16 18 although these issues are still amatterofdebate. 16,17 The current large-size cohort enabled us to further analyze the impact of TBI among the subgroups with these risk factors for graft rejection in the CBT cohort. This study uniquely demonstrated that the addition of TBI to conditioning could provide an advantage for increased neutrophil engraftment in CBT with HLAdisparity and anti-hla antibodies. In the case of an immune interaction between the donor and recipient, TBI might have immunosuppressive potential and prevent graft rejection. Conversely, with a lower infused cell dose, the favorable effect seen with the use of TBI might be limited in the specific MAC setting. Further advances will be required to overcome the disadvantage of a lower infused cell dose in CBT. Recently, novel no-tbi regimens like Flu-Bu-Mel have been investigated and seem to be promising for achieving sufficient neutrophil engraftment in CBT recipients with high-risk leukemia. 22 Although the number of patients in the current cohort who received Flu-Bu-Mel was too small to analyze, it might replace TBI in CBT in the future. If we focused on CBT patients who received mycophenolate mofetil (MMF) as GVHD prophylaxis, no significant difference in engraftment was observed among the five conditioning types, although the population was small (data not shown). It could be another promising strategy to include MMF if the addition of TBI was not suitable. This study has several limitations due to its retrospective nature. The selection of the conditioning regimens and the use of TBI are usually based on several clinical factors, including age, disease, and disease risk. Therefore, there might be some selection preference or bias due to the various patient backgrounds. As most CBT used single cord blood in this cohort, other registry studies are necessary to address whether the impact of TBI differ between single and double CBT. However, the use of a large cohort revealed that TBI had different impacts on neutrophil engraftment according to the donor types, and made it possible to identify which recipients may benefit from the use of TBI in CBT. In conclusion, TBI-regimens had no impact on neutrophil engraftment in the current practice of unrelated BMT from MUD or MMUD. However, TBI is still necessary to enhance engraftment in CBT, regardless of the conditioning intensity. Especially, recipients who receive CBT with two or more HLA-allele mismatches or with anti-hla antibodies may benefit from the addition of TBI to the conditioning regimen. Further observational analyses from other large registry databases would be warranted before we can reach definitive conclusions regarding the use of TBI. In addition, even if we used TBI, the probability of neutrophil engraftment was lower in CBT compared with the other BMT. Therefore, further investigation to enhance overall engraftment achievement in CBT is still equired. ACKNOWLEDGMENTS The authors are grateful for the work of all of the physicians and data managers at the centers that contributed valuable data on transplantation to the Japan Society for Hematopoietic Cell Transplantation (JSHCT), the Japan Marrow Donor Program (JMDP), and the Japan Cord Blood Bank Network (JCBBN). They would also like to thank all of the members of the Transplant Registry Unified Management committees at JSHCT, JMDP, and JCBBN for their dedicated data management. This work was supported in part by a Research Grant for Allergic Disease and Immunology from the Japanese Ministry of Health, Labor, and Welfare (Y.A.). H.N. was a recipient of research awards from Jichi Medical University. AUTHOR CONTRIBUTIONS H.N. designed the study, analyzed data, and wrote the manuscript. S.F., K.Y., M.Onizuka., and A.S. advised on methods and wrote the manuscript. K.O., K.M., and N.U. collected data and revised the manuscript. M.T., T.I. and T.F. collected data, revised the manuscript, and were responsible for data management at JMDP, JCBBN, and JSHCT. Y.A. managed the unified registry database and revised the manuscript. M.Ogata. advised on the methods, wrote the manuscript, and was responsible for the Complication-WG of the JSHCT. REFERENCES [1] Gyurkocza B, Sandmaier BM. Conditioning regimens for hematopoietic cell transplantation: one size does not fit all. Blood. 2014;124: 344 353. [2] Nakasone H, Fukuda T, Kanda J, et al. Impact of conditioning intensity and TBI on acute GVHD after hematopoietic cell transplantation. Bone Marrow Transplant. 2015;50:559 565. [3] Rapoport AP, Miller Watelet LF, Linder T, et al. Analysis of factors that correlate with mucositis in recipients of autologous and allogeneic stem-cell transplants. J Clin Oncol. 1999;17:2446 2453. [4] Hill GR, Ferrara JL. The primacy of the gastrointestinal tract as a target organ of acute graft-versus-host disease: rationale for the use of cytokine shields in allogeneic bone marrow transplantation. Blood. 2000;95:2754 2759. [5] Hill GR, Crawford JM, Cooke KR, Brinson YS, Pan L, Ferrara JL. Total body irradiation and acute graft-versus-host disease: the role of gastrointestinal damage and inflammatory cytokines. Blood. 1997; 90:3204 3213. [6] Clift RA, Buckner CD, Appelbaum FR, et al. Allogeneic marrow transplantation in patients with acute myeloid leukemia in first remission: a randomized trial of two irradiation regimens. Blood. 1990;76:1867 1871. [7] Topolsky D, Crilley P, Styler MJ, Bulova S, Brodsky I, Marks DI. Unrelated donor bone marrow transplantation without T cell depletion using a chemotherapy only condition regimen. Low incidence of failed engraftment and severe acute GVHD. Bone Marrow Transplant. 1996;17:549 554. [8] Mehta J, Powles RL, Mitchell P, Rege K, De Lord C, Treleaven J. Graft failure after bone marrow transplantation from unrelated donors using busulphan and cyclophosphamide for conditioning. Bone Marrow Transplant. 1994;13:583 587. [9] Schultz KR, Ratanatharathorn V, Abella E, et al. Graft failure in children receiving HLA-mismatched marrow transplants with busulfan-containing regimens. Bone Marrow Transplant. 1994;13: 817 822.

178 [10] Shaw BE, Russell NH, Devereux S, et al. The impact of donor factors on primary non-engraftment in recipients of reduced intensity conditioned transplants from unrelated donors. Haematologica. 2005;90:1562 1569. [11] Giralt S, Ballen K, Rizzo D, et al. Reduced-intensity conditioning regimen workshop: defining the dose spectrum. Report of a workshop convened by the center for international blood and marrow transplant research. Biol Blood Marrow Transplant. 2009;15:367 369. [12] Marks DI, Wang T, Perez WS, et al. The outcome of full-intensity and reduced-intensity conditioning matched sibling or unrelated donor transplantation in adults with Philadelphia chromosomenegative acute lymphoblastic leukemia in first and second complete remission. Blood. 2010;116:366 374. [13] Luger SM, Ringden O, Zhang MJ, et al. Similar outcomes using myeloablative vs reduced-intensity allogeneic transplant preparative regimens for AML or MDS. Bone Marrow Transplant. 2012;47:203 211. [14] Atsuta Y, Suzuki R, Yoshimi A, et al. Unification of hematopoietic stem cell transplantation registries in Japan and establishment of the TRUMP System. Int J Hematol. 2007;86:269 274. [15] Kanda Y. Investigation of the freely available easy-to-use software EZR for medical statistics. Bone Marrow Transplant. 2013;48:452 458. [16] Ruggeri A. Alternative donors: cord blood for adults. Semin Hematol. 2016;53:65 73. [17] Ruggeri A, Paviglianiti A, Gluckman E, Rocha V. Impact of HLA in cord blood transplantation outcomes. HLA. 2016;87:413 421. [18] Takanashi M, Atsuta Y, Fujiwara K, et al. The impact of anti-hla antibodies on unrelated cord blood transplantations. Blood. 2010; 116:2839 2846. [19] Adkins DR, DiPersio JF. Total body irradiation before an allogeneic stem cell transplantation: is there a magic dose? Curr Opin Hematol. 2008;15:555 560. [20] Wilke C, Holtan SG, Sharkey L, et al. Marrow damage and hematopoietic recovery following allogeneic bone marrow transplantation for acute leukemias: Effect of radiation dose and conditioning regimen. Radiother Oncol. 2016;118:65 71. [21] Stevens CE, Carrier C, Carpenter C, Sung D, Scaradavou A. HLA mismatch direction in cord blood transplantation: impact on outcome and implications for cord blood unit selection. Blood. 2011; 118:3969 3978. [22] Yamamoto H, Uchida N, Yuasa M, et al. A novel reduced-toxicity myeloablative conditioning regimen using full-dose busulfan, fludarabine, and melphalan for single cord blood transplantation provides durable engraftment and remission in nonremission myeloid malignancies. Biol Blood Marrow Transplant. 2016;22:1844 1850. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article.