- Email: [email protected]
Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation
Time-varying effects of graft type in allo-HCT
Journal Pre-proof
Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation Masamitsu Yanada , Takaaki Konuma , Satoshi Yamasaki , Yachiyo Kuwatsuka , Masayoshi Masuko , Masatsugu Tanaka , Yukiyasu Ozawa , Takashi Toya , Takahiro Fukuda , Shuichi Ota , Masashi Sawa , Naoyuki Uchida , Hirohisa Nakamae , Tetsuya Eto , Junya Kanda , Minoko Takanashi , Yoshinobu Kanda , Yoshiko Atsuta , Shingo Yano PII: DOI: Reference:
S1083-8791(19)30656-1 https://doi.org/10.1016/j.bbmt.2019.09.036 YBBMT 55749
To appear in:
Biology of Blood and Marrow Transplantation
Received date: Accepted date:
2 July 2019 27 September 2019
Please cite this article as: Masamitsu Yanada , Takaaki Konuma , Satoshi Yamasaki , Yachiyo Kuwatsuka , Masayoshi Masuko , Masatsugu Tanaka , Yukiyasu Ozawa , Takashi Toya , Takahiro Fukuda , Shuichi Ota , Masashi Sawa , Naoyuki Uchida , Hirohisa Nakamae , Tetsuya Eto , Junya Kanda , Minoko Takanashi , Yoshinobu Kanda , Yoshiko Atsuta , Shingo Yano , Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation, Biology of Blood and Marrow Transplantation (2019), doi: https://doi.org/10.1016/j.bbmt.2019.09.036
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of the American Society for Transplantation and Cellular Therapy
HIGHLIGHTS
This study aimed to investigate time-varying effects of graft type on allogeneic HCT outcomes for AML.
RFS and OS were worse for related PBSCT and UCBT than for related BMT.
Adverse impact of UCBT was observed only during the early phase of transplantation.
Adverse impact of related PBSCT continued even after two years post-transplant.
Our findings raise concerns regarding the increased risk of late non-relapse mortality with the use of PBSC grafts.
1
Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation
Masamitsu Yanada1, Takaaki Konuma2, Satoshi Yamasaki3, Yachiyo Kuwatsuka4, Masayoshi Masuko5, Masatsugu Tanaka6, Yukiyasu Ozawa7, Takashi Toya8, Takahiro Fukuda9, Shuichi Ota10, Masashi Sawa11, Naoyuki Uchida12, Hirohisa Nakamae13, Tetsuya Eto14, Junya Kanda15, Minoko Takanashi16, Yoshinobu Kanda17, Yoshiko Atsuta18,19, Shingo Yano20
1
Aichi Cancer Center, Nagoya, Japan; 2The Institute of Medical Science, The
University of Tokyo, Tokyo, Japan; 3National Hospital Organization Kyusyu Medical Center, Fukuoka, Japan;4Nagoya University Hospital, Nagoya, Japan; 5
Niigata University Medical and Dental Hospital, Niigata, Japan; 6Kanagawa
Cancer Center, Yokohama, Japan; 7Japanese Red Cross Nagoya First Hospital, Nagoya, Japan; 8Tokyo Metropolitan Komagome Hospital, Tokyo, Japan; 9
National Cancer Center Hospital, Tokyo, Japan; 10Sapporo Hokuyu Hospital,
Sapporo, Japan; 11Anjo Kosei Hospital, Anjo, Japan;
12
Toranomon Hospital,
Tokyo, Japan; 13Graduate School of Medicine, Osaka City University, Osaka, Japan; 14Hamanomachi Hospital, Fukuoka, Japan;
15
Graduate School of
Medicine, Kyoto University, Kyoto, Japan; 16Japanese Red Cross Society, Tokyo, Japan; 17Jichi Medical University, Shimotsuke, Japan;
18
for Hematopoietic Cell Transplantation, Nagoya, Japan;
Japanese Data Center 19
Nagoya University
Graduate School of Medicine, Nagoya, Japan; 20The Jikei University School of Medicine, Tokyo, Japan
2
Corresponding author: Masamitsu Yanada, MD, PhD Address: Department of Hematology and Cell Therapy, Aichi Cancer Center, 1-1 Kanokoden, Chikusa-ku, Nagoya, 464-8681, Japan TEL: +81-52-762-6111 FAX: +81-52-764-2967 E-mail: [email protected]
Short title: Time-varying effects of graft type in allo-HCT
Keywords: acute myeloid leukemia, allogeneic hematopoietic cell transplantation, bone marrow transplantation, peripheral blood stem cell transplantation, umbilical cord blood transplantation, late effect
Financial disclosure: This work was supported in part by a grant from the Practical Research Project for Allergic Diseases and Immunology (Research Technology of Medical Transplantation) from the Japan Agency for Medical Research and Development (AMED), grant number: 18ek0510023h0002, and a grant from the Aichi Cancer Research Foundation.
Conflict of interest statement: There are no conflicts of interest to report.
ABSTRACT
3
This study aimed to investigate time-varying effects of graft type on outcomes for patients with acute myeloid leukemia (AML) undergoing allogeneic hematopoietic cell transplantation. For this purpose, we analyzed 3952 patients comprising 720 of whom underwent matched related bone marrow transplantation (BMT), 1004 matched related peripheral blood stem cell transplantation (PBSCT), 856 matched unrelated BMT, and 1372 umbilical cord blood transplantation (UCBT). The 4-year relapse-free survival (RFS) rates were 59.1%, 52.8%, 59.5%, and 50.6%, respectively. Compared to related BMT, related PBSCT, unrelated BMT and UCBT were associated with higher risk of non-relapse mortality, and unrelated BMT and UCBT with lower risk of relapse. As a result, both RFS and overall survival (OS) were comparable between related BMT and unrelated BMT, but were worse for related PBSCT and UCBT than for related BMT. Adverse impact of UCBT was observed only during the early phase of transplantation, whereas that of related PBSCT continued even after two years post-transplant. Our findings raise concerns about the increased risk of late non-relapse mortality with the use of PBSC grafts, and suggest that related BMT is preferable to related PBSCT, and matched unrelated BMT is the next choice in the absence of a matched related donor.
4
INTRODUCTION Allogeneic hematopoietic cell transplantation (HCT) is an effective treatment that makes long-term survival possible for patients with high-risk acute myeloid leukemia (AML) [1-3]. Because of the highly complex nature of the modality, however, success depends on multiple factors. The graft type, i.e., related and unrelated bone marrow (BM), peripheral blood stem cells (PBSC), or umbilical cord blood (UCB), is among the most significant transplantation-related factors, and has been extensively investigated in a number of prospective and retrospective studies [4-23]. The Cox proportional hazards model is widely used in survival analysis to assess the prognostic impact of individual covariates [24]. This method is reliable assuming that proportionality of the hazards does not change with time. However, the effect of graft type on transplantation outcomes may be time-dependent, because the graft type affects a variety of outcome events including engraftment, graft-versus-host disease (GVHD), and relapse due to the graft-versus-leukemia (GVL) effect, each of which occurs at different time points after transplantation. This study aimed to investigate time-varying effects of graft type on outcomes for patients with AML undergoing allogeneic HCT. To this end, we used landmark analysis and additive hazards model analysis, the former for evaluating late effects [24], and the latter for estimating how the effect of covariates changes over time [25].
5
METHODS Study Population All data were collected through the Transplant Registry Unified Management Program sponsored by the Japanese Society for Hematopoietic Cell Transplantation and the Japanese Data Center for Hematopoietic Cell Transplantation. This registration program covers nearly all transplantation centers nationwide, and each participating center is required to report annually follow-up information for consecutively registered HCTs [26-28]. Patients with AML (except acute promyelocytic leukemia) were selected from the database if they were aged 16 years or older, were in first or second complete remission (CR) at the time of transplantation, and had undergone their first allogeneic HCT between 2000 and 2016 from a matched sibling donor, a matched unrelated donor, or a UCB unit that met the criteria specified below. For a matched sibling donor, HLA-A, -B, and -DR at the antigen level were taken into consideration, as were HLA-A, -B, -C and -DRB1 at the allele level for a matched unrelated donor. The UCB unit had to be a single-unit containing a total nucleated cell (TNC) dose 2.0 × 107/kg of the recipient’s body weight and show at least 4/6-match at the antigen level for HLA-A, -B, and -DR, in line with clinical practice in Japan [29]. Patients receiving PBSC from an unrelated donor were not included, because unrelated PBSC transplantation (PBSCT) was rarely performed in Japan during the study period [30]. The study presented here
6
therefore compared four graft types: related BM, related PBSC, unrelated BM, and UCB. This study was approved by the Institutional Review Board of the Aichi Cancer Center, and was conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from each patient.
Definitions Cytogenetic risk was classified in accordance with the criteria specified by the National Comprehensive Cancer Network Guidelines as described in detail elsewhere [31]. The conditioning intensity was determined in line with consensus criteria [32]. Neutrophil and platelet engraftment was defined as the first of three consecutive measures with a neutrophil count 0.5 × 109/L and a platelet count 50 × 109/L. Acute and chronic GVHD were assessed according to the standard criteria [33,34]. For the analysis of chronic GVHD, only patients who had survived 100 days post-transplant without relapse were included.
Statistical Analysis The primary objective of this study was to assess the effect of graft type on relapse-free survival (RFS), and the secondary objective was to assess such effects on overall survival (OS), relapse, and non-relapse mortality. RFS was defined as the time from transplantation to relapse, death or last visit, and OS as the time from transplantation to death or last visit. The probabilities of RFS and
7
OS were estimated by means of the Kaplan–Meier estimator. Multivariate analyses were performed using the Cox proportional hazards model, and hazard ratios (HRs) were calculated together with the corresponding 95% confidence intervals (CIs). The final multivariate model was built using a backward stepwise model selection approach, with a significance cutoff level of 0.05 for removing factors from the model. The variable for graft type, as the main effect, was entered into the final model irrespective of its statistical significance. Relapse and non-relapse mortality were considered to be mutually competing events, and their incidence was estimated with the aid of the cumulative incidence estimator. The Fine and Gray model was used for multivariate analyses. Cumulative incidence of neutrophil and platelet engraftment, and GVHD was also calculated by accommodating competing risks (death for neutrophil and platelet engraftment, and death and relapse for GVHD). Landmark analysis was performed to assess the impact of prognostic variables on RFS, OS, relapse and non-relapse mortality for patients alive without relapse 2 years post transplantation. The additive hazards model proposed by Aalen was used for analysis of time-dependent changes in the effect of graft type on RFS [25]. All statistical analyses were performed with Stata version 14.2 software (StataCorp. LP, College Station, TX, USA).
RESULTS Patients
8
A total of 3952 patients were eligible for analysis, of whom 720 underwent related BM transplantation (BMT), 1004 related PBSCT, 856 unrelated BMT, and 1372 UCB transplantation (UCBT). The median follow-up for surviving patients was 4.4 years (0.1–17.3 years). Table 1 summarizes patient characteristics according to graft type. Several differences were noted among the four groups. Patients undergoing HCT from a related donor were more likely to be transplanted during first CR (78% for related BMT, 79% for related PBSCT, 71% for unrelated BMT, and 69% for UCBT). UCBT recipients were older and received reduced-intensity conditioning more frequently. The median age of patients was 43 years for related BMT, 44 for related PBSCT, 46 for unrelated BMT, and 51 for and UCBT, and reduced-intensity conditioning was used for 19%, 25%, 21%, and 36%, respectively. Transplantation from a female donor to a male recipient was less common for unrelated BMT (26% for related BMT, 30% for related PBSCT, 14% for unrelated BMT, and 29% for UCBT). The percentage of patients transplanted in the early years covered by this study (2000–2010 vs 2011–2016, where the median was used as a cutoff) was 63% for related BMT, 51% for related PBSCT, 58% for unrelated BMT, and 39% for UCBT. For GVHD prophylaxis, anti-thymocyte globulin (ATG) was used in only 1.8% of the overall study population. Owing to the limited number, the use of ATG was not considered for further analyses.
Engraftment and GVHD
9
Table 2 summarizes cumulative incidences of engraftment and GVHD according to graft type, and Table 3 adjusted effects of graft type using related BMT as the reference. In multivariate analysis, the likelihood of neutrophil and platelet engraftment was higher for related PBSCT, while it was lower for UCBT. Unrelated BMT showed a similar likelihood of neutrophil engraftment, but a lower likelihood of platelet engraftment. Related PBSCT was associated with higher risk of grade III–IV acute GVHD as well as overall and extensive chronic GVHD. Unrelated BMT was at higher risk of grade II–IV acute GVHD, whereas the risks of grade III–IV acute GVHD and chronic GVHD were not significantly different from those for related BMT. UCBT was associated with higher risk of grade II–IV and grade III–IV acute GVHD, but the risk of extensive chronic GVHD was lower conversely. Full results of the multivariate analysis are provided in Supplementary Table 1.
Relapse and Non-Relapse Mortality Cumulative incidences of relapse and non-relapse mortality according to graft type are summarized in Table 2 and graphically displayed in Figures 1A and 1B. The cumulative incidence of relapse was similar for related BMT and related PBSCT, but lower for unrelated BMT and UCBT (27.8%, 28.3%, 21.1%, 18.8% at 4 years, respectively; Figure 1A). The cumulative incidence of non-relapse mortality was highest for UCBT and lowest for related BMT, with 12.9% for related BMT, 18.9% for related PBSCT, 19.3% for unrelated BMT, and 30.6% for
10
UCBT (Figure 1B). Multivariate analysis showed that, compared to related BMT, related PBSCT was associated with equivalent risk of relapse and higher risk of non-relapse mortality, unrelated BMT with lower risk of relapse and higher risk of non-relapse mortality, and UCBT with lower risk of relapse and higher risk of non-relapse mortality (Table 3). Full results of the multivariate analysis are provided in Supplementary Table 1.
Relapse-Free and Overall Survival The probabilities of RFS and OS according to graft type are summarized in Table 2 and graphically displayed in Figures 1C and 1D. The RFS curves were identical for related BMT and unrelated BMT, but lower for related PBSCT and UCBT (Figure 1C). The 4-year RFS rates were 59.1% for related BMT, 52.8% for related PBSCT, 59.5% for unrelated BMT, and 50.6% for UCBT. Similar trends were observed for OS (Figure 1D), with corresponding rates of 64.5%, 58.2%, 62.7%, and 53.9% at 4 years. Multivariate analysis using related BMT as the reference showed that both RFS and OS were inferior for related PBSCT and UCBT, and were comparable for unrelated BMT (Table 3). Full results of the multivariate analysis are provided in Supplementary Table 1. Effects of several covariates on RFS, the primary endpoint of this study, were assessed within individual graft type groups. RFS was not affected by year of transplantation (1992–2010 or 2011–2016) for each group (P = 0.180 for related BMT, P = 0.714 for related PBSCT, P = 0.752 for unrelated BMT, and P = 0.666 for UCBT).
11
As for unrelated BMT, although the donor age was significantly higher by 4 years in the later cohort (P < 0.001), there was no difference in RFS between patients whose donor age was below or above the median value of 36 years (P = 0.140). For patients undergoing UCBT, RFS did not differ by whether they received a UCB unit containing a TNC dose below or above the median value of 2.67 × 107/kg of the recipient’s body weight (P = 0.199).
Outcomes after a 2-Year Landmark Period Landmark analysis was performed for assessment of the late effects of graft type on outcomes. The landmark was set at 2 years after transplantation, and the 1946 patients who were alive without relapse at this time point were analyzed. The median follow-up for surviving patients was 3.5 years (0.0–15.3 years) after the landmark. The cumulative incidence of relapse was 3.2% for related BMT, 7.5% for related PBSCT, 4.0% for unrelated BMT, and 3.9% for UCBT at 3 years from the 2-year landmark (Figure 2A). There was no difference in the risk of relapse between related BMT and other transplantations in multivariate analysis (Table 4). The cumulative incidence of non-relapse mortality was 7.1% for related BMT, 10.7% for related PBSCT, 3.2% for unrelated BMT, and 7.7% for UCBT (Figure 2B) at 3 years from the 2-year landmark. Multivariate analysis showed that only related PBSCT significantly correlated with higher risk of non-relapse mortality compared to related BMT (Table 4). The RFS and OS rates at 3 years after the 2-year landmark were
12
89.7% and 90.7% for related BMT, 81.8% and 85.4% for related PBSCT, 92.8% and 94.3% for unrelated BMT, and 88.5% and 90.9% for UCBT, respectively (Figures 2C and 2D). Multivariate analysis showed that only related PBSCT was associated with worse RFS and OS than related BMT, while unrelated BMT and UCBT had comparable RFS and OS with related BMT. Full results of the multivariate analysis are provided in Supplementary Table 2.
Changes in the Effect of Graft Type on RFS We next used the additive hazards model to examine how the effect of graft type on RFS changes over time. Figure 3 shows the cumulative coefficients of each graft type versus related BMT. The plots display changes in hazard rates over time, with an ascending slope indicating increased hazard rates, a descending slope reduced hazard rates, and a flat slope no time-varying effect. The results have been adjusted for factors associated with RFS, i.e., age, sex, performance status and cytogenetic risk. The curve for related PBSCT fluctuated for the first 1 year, and continued to rise thereafter (Figure 3A), indicating that the risk for related PBSCT increased steadily after 1 year. The curve for unrelated BMT increased for the first half year, then decreased, and had a roughly flat slope until 4 years, after which it decreased further (Figure 3B). This indicates that the risk for unrelated BMT increased in the early period of around 6 months post-transplantation, but decreased after that. The curve for UCBT increased steeply for the first few months and then increased gradually for the
13
first one and a half years. The slope then changed from positive to negative, and became flat thereafter (Figure 3C). This suggests that the risk for UCBT increased substantially in the early period, especially for the first few months, but there was no further increase in the risk in the late period.
DISCUSSION The study presented here show that graft type significantly affected engraftment, GVHD, relapse and non-relapse mortality, which collectively impacted survival outcomes in a time-dependent manner. Adverse impacts of unrelated BMT and UCBT were observed only during the early phase of transplantation, whereas that of related PBSCT continued even after two years post-transplant. PBSC has now largely replaced BM as stem cell source for allogeneic HCT. Although its use for unrelated HCT is still very limited in Japan because unrelated PBSCT has been used in clinical practice since 2010 [30], the annual number is expected to increase. Owing to the inclusion of more CD34+ cells and T-cells in the graft, PBSCT is characterized by faster engraftment but also higher risk of GVHD compared to BMT. While GVHD is a major complication associated with allogeneic HCT, it is accompanied by a potentially protective effect against relapse, and a few randomized studies have shown a reduction in relapse with PBSCT versus BMT [4,17]. Our analysis, which focused on patients with AML in CR, shows that the use of related PBSCT increased non-relapse
14
mortality without a reduction in relapse, leading to worse RFS and OS compared to related BMT. Furthermore, the adverse impact of related PBSCT on non-relapse mortality, RFS and OS remained even after 2 years post-transplant, with the cumulative hazard curve for RFS ascending steadily thereafter. Related PBSCT was also associated with significantly higher risk of chronic GVHD compared to related BMT. These results taken together suggest that higher risk of chronic GVHD for related PBSCT resulted in later mortality but did not help prevent relapse for our patients. In addition, related PBSCT was associated with increased risk of grade III–IV acute GVHD. Despite the lack of data on unrelated PBSCT for this study, it is conceivable that higher risk of GVHD is a matter of particular concern when choosing unrelated PBSCT instead of unrelated BMT as previously reported [18]. Since only approximately 40% of transplantation candidates have a matched related donor [19,35,36], an alternative donor needs to be found for the remaining 60% to undergo allogeneic HCT. With the aid of high-resolution HLA typing, there have been significant improvements in the outcome of unrelated HCT [37,38]. Several studies have shown that allogeneic HCT from an unrelated donor matched at the allele level yields results similar to those of allogeneic HCT from a matched related donor [10,13-15], and our findings are consistent with those previously reported. A higher risk of non-relapse mortality for our patients undergoing unrelated BMT was counterbalanced by reduced risk of relapse, resulting in RFS and OS similar to those for related BMT. Although the risk of
15
grade II–IV acute GVHD was higher for unrelated BMT, there was no significant increase in the risk of grade III–IV acute GVHD or chronic GVHD. The cumulative hazard curve on RFS peaked around the first half year, and no late effect was observed. These results support the notion that a matched unrelated donor is a reasonable alternative both in the short- and long-term when a matched related donor is not available. Here, attention should be paid to the fact that the longer time needed for unrelated donor search sometimes precludes performing a transplantation during CR. On this point, our previous decision analysis showed that unrelated BMT is preferable to immediate UCBT, if unrelated BMT can be performed within 3–9 months after achieving first CR [20]. UCB is another stem cell source for allogeneic HCT. In addition to its immediate availability, UCB has the benefit of greater tolerance of HLA disparity. On the other hand, the low number of hematopoietic stem cells contained in a UCB unit constitutes a major limitation to its use, which can lead to delayed engraftment, higher risk of post-transplant infection, and early death. Our data show that UCBT had a more potent protective effect against relapse than related BMT, but this advantage was outweighed by higher risk of non-relapse mortality, leading to significantly worse RFS and OS. While at the same time, the risk of extensive chronic GVHD was lower, and there was no late adverse effect on survival outcomes. The cumulative hazard curve for RFS peaked at one and half years despite a steep increase soon after transplantation. These findings indicate that the adverse impact of UCBT is limited mostly to the early phase of
16
transplantation. One may argue that our criterion for UCB unit selection of TNC dose 2.0 × 107/kg was less stringent than those used in Western countries, which allowed for an inclusion of patients who received TNC doses that were considered inadequate. In this regard, however, we had previously shown that the TNC dose in excess of the minimum requirement of 2.0 × 107/kg were not predictive for OS, relapse, or non-relapse mortality [29]. Moreover, our finding has been recently corroborated by a joint study by the European Society for Blood and Marrow Transplantation and the Japanese Society of Hematopoietic Cell Transplantation, in which the TNC dose showed no significant effect on OS [39]. At any rate, our data highlight that the reduction in non-relapse mortality during the early phase of transplantation is the key to the success of UCBT. Several investigators have implemented strategies, including double-unit UCBT [40] and ex vivo expansion of cord blood cells [41]. These efforts, in concert with the establishment of better unit selection criteria as well as improvements in supportive care measures, are expected to overcome the current limitations of UCBT. Selection bias must be considered when interpreting our results. Which patients received which grafts was the result of non-randomized selection, and we do not know, for example, whether patients who underwent UCBT did so as planned initially or only after the failure of an unrelated donor search. It is also possible that our results were confounded by changes in practice implemented during the study period, including patient selection, expansion of the unrelated
17
donor registry, the accumulated experience of each transplantation center, and advances in transplantation procedures and supportive care measures. Although the large number of patients assessed in our study allowed for accurate multivariate analysis by adjusting for the confounding effects of known factors, such unknown or unmeasured factors may have affected the results. This type of bias can theoretically be resolved by a prospective randomized study; however, conducting such a study is unfeasible. This situation enables retrospective studies such as ours to serve as important sources of information. While acknowledging potential limitations inherent to a registry-based study, an inclusion of a homogeneous patient population, i.e., adult patients with AML in first or second CR, represents the unique strength of this study. Additional caution needs to be exercised in that our findings were based exclusively on Japanese patient data. Because racial differences in the risk of GVHD and thus possibly of GVL effect are well recognized [42,43], it may be premature to apply our conclusions to other patient populations. Finally, although our registry data did not contain information on costs or the quality of life (QOL), it is conceivable that the graft type may affect these outcomes. Late complications such as chronic GVHD have been shown to deteriorate patients’ QOL [44,45], and increase medical costs [46,47]. Addressing how the graft type impacts these outcomes will provide additional important insights into graft selection. In conclusion, this study presents the time-varying effects of graft type on outcomes after allogeneic HCT. Our results raise a particular concern about the
18
increased risk of late non-relapse mortality with the use of PBSC grafts, and suggest that related BMT is preferable to related PBSCT, and BMT from a matched unrelated donor is the next choice for patients with AML in CR who need allogeneic HCT but lack a matched related donor. Future studies that deal with allogeneic HCT using various types of grafts should give careful consideration to the proportional hazards assumption.
REFERENCES 1.
Gupta V, Tallman MS, Weisdorf DJ. Allogeneic hematopoietic cell transplantation for adults with acute myeloid leukemia: myths, controversies, and unknowns. Blood. 2011;117:2307-2318.
2.
Yanada M. Allogeneic hematopoietic cell transplantation for acute myeloid leukemia during first complete remission: a clinical perspective. Int J Hematol. 2015;101:243-254.
3.
Cornelissen JJ, Blaise D. Hematopoietic stem cell transplantation for patients with AML in first complete remission. Blood. 2016;127:62-70.
4.
Powles R, Mehta J, Kulkarni S, et al. Allogeneic blood and bone-marrow stem-cell transplantation in haematological malignant diseases: a randomised trial. Lancet. 2000;355:1231-1237.
5.
Vigorito AC, Marques Junior JF, Aranha FJ, Oliveira GB, Miranda EC, De Souza CA. A randomized, prospective comparison of allogeneic bone marrow and peripheral blood progenitor cell transplantation in the treatment of hematologic malignancies: an update. Haematologica. 2001;86:665-666.
6.
Couban S, Simpson DR, Barnett MJ, et al. A randomized multicenter comparison of bone marrow and peripheral blood in recipients of matched sibling allogeneic transplants for myeloid malignancies. Blood.
19
2002;100:1525-1531. 7.
Laughlin MJ, Eapen M, Rubinstein P, et al. Outcomes after transplantation of cord blood or bone marrow from unrelated donors in adults with leukemia. N Engl J Med. 2004;351:2265-2275.
8.
Rocha V, Labopin M, Sanz G, et al. Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. N Engl J Med. 2004;351:2276-2285.
9.
Moore J, Nivison-Smith I, Goh K, et al. Equivalent survival for sibling and unrelated donor allogeneic stem cell transplantation for acute myelogenous leukemia. Biol Blood Marrow Transplant. 2007;13:601-607.
10.
Schetelig J, Bornhauser M, Schmid C, et al. Matched unrelated or matched sibling donors result in comparable survival after allogeneic stem-cell transplantation in elderly patients with acute myeloid leukemia: a report from the cooperative German Transplant Study Group. J Clin Oncol. 2008;26:5183-5191.
11.
Atsuta Y, Suzuki R, Nagamura-Inoue T, et al. Disease-specific analyses of unrelated cord blood transplantation compared with unrelated bone marrow transplantation in adult patients with acute leukemia. Blood. 2009;113:1631-1638.
12.
Nagafuji K, Matsuo K, Teshima T, et al. Peripheral blood stem cell versus bone marrow transplantation from HLA-identical sibling donors in patients with leukemia: a propensity score-based comparison from the Japan Society for Hematopoietic Stem Cell Transplantation registry. Int J Hematol. 2010;91:855-864.
13.
Walter RB, Pagel JM, Gooley TA, et al. Comparison of matched unrelated and matched related donor myeloablative hematopoietic cell transplantation for adults with acute myeloid leukemia in first remission. Leukemia. 2010;24:1276-1282.
14.
Gupta V, Tallman MS, He W, et al. Comparable survival after HLA-well-matched unrelated or matched sibling donor transplantation for acute myeloid leukemia in first remission with unfavorable cytogenetics at diagnosis. Blood. 2010;116:1839-1848.
20
15.
Schlenk RF, Dohner K, Mack S, et al. Prospective evaluation of allogeneic hematopoietic stem-cell transplantation from matched related and matched unrelated donors in younger adults with high-risk acute myeloid leukemia: German-Austrian trial AMLHD98A. J Clin Oncol. 2010;28:4642-4648.
16.
Friedrichs B, Tichelli A, Bacigalupo A, et al. Long-term outcome and late effects in patients transplanted with mobilised blood or bone marrow: a randomised trial. Lancet Oncol. 2010;11:331-338.
17.
Mielcarek M, Storer B, Martin PJ, et al. Long-term outcomes after transplantation of HLA-identical related G-CSF-mobilized peripheral blood mononuclear cells versus bone marrow. Blood. 2012;119:2675-2678.
18.
Anasetti C, Logan BR, Lee SJ, et al. Peripheral-blood stem cells versus bone marrow from unrelated donors. N Engl J Med. 2012;367:1487-1496.
19.
Yanada M, Kurosawa S, Yamaguchi T, et al. Effect of related donor availability on outcome of AML in the context of related and unrelated hematopoietic cell transplantation. Bone Marrow Transplant. 2013;48:390-395.
20.
Yanada M, Kanda J, Ohtake S, et al. Unrelated bone marrow transplantation or immediate umbilical cord blood transplantation for patients with acute myeloid leukemia in first complete remission. Eur J Haematol. 2016;97:278-287.
21.
Terakura S, Atsuta Y, Tsukada N, et al. Comparison of outcomes of 8/8 and 7/8 allele-matched unrelated bone marrow transplantation and single-unit cord blood transplantation in adults with acute leukemia. Biol Blood Marrow Transplant. 2016;22:330-338.
22.
Savani BN, Labopin M, Blaise D, et al. Peripheral blood stem cell graft compared to bone marrow after reduced intensity conditioning regimens for acute leukemia: a report from the ALWP of the EBMT. Haematologica. 2016;101:256-262.
23.
Versluis J, Labopin M, Ruggeri A, et al. Alternative donors for allogeneic hematopoietic stem cell transplantation in poor-risk AML in CR1. Blood
21
Adv. 2017;1:477-485. 24.
Delgado J, Pereira A, Villamor N, Lopez-Guillermo A, Rozman C. Survival analysis in hematologic malignancies: recommendations for clinicians. Haematologica. 2014;99:1410-1420.
25.
Aalen OO. A linear regression model for the analysis of life times. Stat Med. 1989;8:907-925.
26.
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.
27.
Atsuta Y. Introduction of Transplant Registry Unified Management Program 2 (TRUMP2): scripts for TRUMP data analyses, part I (variables other than HLA-related data). Int J Hematol. 2016;103:3-10.
28.
Kanda J. Scripts for TRUMP data analyses. Part II (HLA-related data): statistical analyses specific for hematopoietic stem cell transplantation. Int J Hematol. 2016;103:11-19.
29.
Yanada M, Konuma T, Kuwatsuka Y, et al. Unit selection for umbilical cord blood transplantation for adults with acute myeloid leukemia in complete remission: a Japanese experience. Bone Marrow Transplant. 2019 May 8. [Epub ahead of print].
30.
Saito H, Ito M, Kato S, et al. The Japan Marrow Donor Program, 25 years of experience in achieving 20,000 bone marrow transplantations: organization structure, activity, and financial basis. Bone Marrow Transplant. 2018;53:609-616.
31.
Yanada M, Mori J, Aoki J, et al. Effect of cytogenetic risk status on outcomes for patients with acute myeloid leukemia undergoing various types of allogeneic hematopoietic cell transplantation: an analysis of 7812 patients. Leuk Lymphoma. 2018;59:601-609.
32.
Giralt S, Logan B, Rizzo D, et al. Reduced-intensity conditioning for unrelated donor progenitor cell transplantation: long-term follow-up of the first 285 reported to the national marrow donor program. Biol Blood Marrow Transplant. 2007;13:844-852.
33.
Sullivan KM, Agura E, Anasetti C, et al. Chronic graft-versus-host disease
22
and other late complications of bone marrow transplantation. Semin Hematol. 1991;28:250-259. 34.
Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15:825-828.
35.
Mawad R, Gooley TA, Sandhu V, et al. Frequency of allogeneic hematopoietic cell transplantation among patients with high- or intermediate-risk acute myeloid leukemia in first complete remission. J Clin Oncol. 2013;31:3883-3888.
36.
Labopin M, Gorin NC, Polge E, et al. A prospective registration study to determine feasibility of hematopoietic SCT in adults with acute leukemia: planning, expectations and reality. Bone Marrow Transplant. 2014;49:376-381.
37.
Majhail NS, Chitphakdithai P, Logan B, et al. Significant improvement in survival after unrelated donor hematopoietic cell transplantation in the recent era. Biol Blood Marrow Transplant. 2015;21:142-150.
38.
Yanada M, Masuko M, Mori J, et al. Patients with acute myeloid leukemia undergoing allogeneic hematopoietic cell transplantation: trends in survival during the past two decades. Bone Marrow Transplant. 2019;54:578-586.
39.
Kanda J, Hayashi H, Ruggeri A, et al. Prognostic factors for adult single cord blood transplantation among European and Japanese populations: the Eurocord/ALWP-EBMT and JSHCT/JDCHCT collaborative study. Leukemia. 2019.
40.
Barker JN, Weisdorf DJ, DeFor TE, et al. Transplantation of 2 partially HLA-matched umbilical cord blood units to enhance engraftment in adults with hematologic malignancy. Blood. 2005;105:1343-1347.
41.
de Lima M, McNiece I, Robinson SN, et al. Cord-blood engraftment with ex vivo mesenchymal-cell coculture. N Engl J Med. 2012;367:2305-2315.
42.
Oh H, Loberiza FR, Jr., Zhang MJ, et al. Comparison of graft-versus-host-disease and survival after HLA-identical sibling bone marrow transplantation in ethnic populations. Blood. 2005;105:1408-1416.
23
43.
Morishima Y, Kawase T, Malkki M, et al. Significance of ethnicity in the risk of acute graft-versus-host disease and leukemia relapse after unrelated donor hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2013;19:1197-1203.
44.
Pidala J, Kurland B, Chai X, et al. Patient-reported quality of life is associated with severity of chronic graft-versus-host disease as measured by NIH criteria: report on baseline data from the Chronic GVHD Consortium. Blood. 2011;117:4651-4657.
45.
Kurosawa S, Oshima K, Yamaguchi T, et al. Quality of Life after Allogeneic Hematopoietic Cell Transplantation According to Affected Organ and Severity of Chronic Graft-versus-Host Disease. Biol Blood Marrow Transplant. 2017;23:1749-1758.
46.
Saito AM, Cutler C, Zahrieh D, et al. Costs of allogeneic hematopoietic cell transplantation with high-dose regimens. Biol Blood Marrow Transplant. 2008;14:197-207.
47.
Khera N, Emmert A, Storer BE, Sandmaier BM, Alyea EP, Lee SJ. Costs of allogeneic hematopoietic cell transplantation using reduced intensity conditioning regimens. Oncologist. 2014;19:639-644.
24
FIGURE LEGENDS
Figure 1. Cumulative incidence of relapse (A), cumulative incidence of non-relapse mortality (B), relapse-free survival (C), and overall survival (D) by graft type. Patients undergoing matched related bone marrow transplantation (BMT; N = 720), matched related peripheral blood stem cell transplantation (PBSCT; N = 1,004), matched unrelated BMT (N = 856), and umbilical cord blood transplantation (UCBT; N = 1,372) are compared.
25
Figure 2. Cumulative incidence of relapse (A), cumulative incidence of non-relapse mortality (B), relapse-free survival (C), and overall survival (D) by graft type in a landmark analysis at 2 years after transplantation. Only those who survived without relapse for at least 2 years are included. Patients undergoing matched related bone marrow transplantation (BMT; N = 384), matched related peripheral blood stem cell transplantation (PBSCT; N = 510), matched unrelated BMT (N = 482), and umbilical cord blood transplantation (UCBT; N = 570) are compared.
26
27
Figure 3. Cumulative coefficients in terms of relapse-free survival (RFS) obtained from the additive hazards model for related peripheral blood stem cell transplantation (PBSCT) (A), unrelated bone marrow transplantation (BMT) (B), and umbilical cord blood transplantation (UCBT) (C), in comparison to related BMT. The plots display changes in hazard rates over time, with an ascending slope indicating increased hazard rates, a descending slope reduced hazard rates, and a flat slope no time-varying effect. Results are adjusted for age, sex, performance status and cytogenetic risk, all of which were identified as factors significantly associated with RFS.
28
Table 1. Characteristics of Patients According to Graft Type Related BMT
Related PBSCT
Unrelated BMT
UCBT
N = 720
N = 1004
N = 856
N = 1372
Age Median (range), years
43
(16–73)
44
(16–73)
46
(16–70)
51
(16–85)
<50 years
483
(67%)
632
(63%)
505
(59%)
633
(46%)
50 years
237
(33%)
372
(37%)
351
(41%)
739
(54%)
Male
396
(55%)
596
(59%)
482
(56%)
705
(51%)
Female
324
(45%)
408
(41%)
374
(44%)
666
(49%)
0–1
695
(97%)
976
(98%)
784
(98%)
1251
(96%)
2
20
(3%)
23
(2%)
19
(2%)
58
(4%)
Myeloablative
543
(81%)
719
(75%)
671
(79%)
876
(64%)
Reduced-intensity
124
(19%)
237
(25%)
175
(21%)
496
(36%)
Favorable
120
(17%)
155
(15%)
136
(16%)
162
(12%)
Intermediate
444
(62%)
645
(64%)
545
(64%)
891
(65%)
Poor
102
(14%)
128
(13%)
101
(12%)
233
(17%)
54
(8%)
76
(8%)
74
(9%)
86
(6%)
CR1
558
(78%)
793
(79%)
610
(71%)
942
(69%)
CR2
162
(23%)
211
(21%)
246
(29%)
430
(31%)
Female to male
186
(26%)
294
(30%)
121
(14%)
327
(29%)
Others
532
(74%)
700
(70%)
732
(86%)
805
(71%)
2000–2010
458
(63%)
511
(51%)
495
(58%)
538
(39%)
2011–2016
262
(36%)
493
(49%)
361
(42%)
834
(69%)
Sex
ECOG performance status
Conditioning
Cytogenetic risk
Unevaluable Disease status
Sex mismatch
Year of transplantation
BMT indicates bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation; ECOG, Eastern Cooperative Oncology Group; CR, complete remission.
29
Table 2. Post-transplant Outcomes by Graft Type
Related
Related
Unrelated
BMT
PBSCT
BMT
95.5%
98.4%
95.9%
78.3%
89.2%
91.5%
86.5%
56.7%
25.6%
28.9%
33.2%
36.5%
6.2%
10.2%
9.2%
10.5%
Chronic GVHD, at 2 years
34.6%
51.5%
38.8%
26.6%
Extensive chronic GVHD,
19.7%
33.0%
23.2%
11.5%
Relapse, at 4 years
27.8%
28.3%
21.1%
18.8%
Non-relapse mortality, at 4
12.9%
18.9%
19.3%
30.6%
59.1%
52.8%
59.5%
50.6%
64.5%
58.2%
62.7%
53.9%
Neutrophil engraftment, at
UCBT
30 days Platelet engraftment, at 60 days Grade II–IV acute GVHD, at 100 days Grade III–IV acute GVHD, at 100 days
at 2 years
years Relapse-free survival, at 4 years Overall survival, at 4 years
BMT indicates bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation; GVHD, graft-versus-host disease.
30
Table 3. Adjusted Effects of Graft Types on Outcomes Estimated by Multivariate Analysis HR (95% CI)
P-value
Neutrophil engraftment Related BMT
1.00
Related PBSCT
1.89 (1.68–2.12)
<0.001
Unrelated BMT
0.95 (0.86–1.05)
0.327
UCBT
0.43 (0.40–0.48)
<0.001
Platelet engraftment Related BMT
1.00
Related PBSCT
1.52 (1.34–1.73)
<0.001
Unrelated BMT
0.82 (0.73–0.92)
0.001
UCBT
0.33 (0.30–0.37)
<0.001
Grade II–IV acute GVHD Related BMT
1.00
Related PBSCT
1.17 (0.97–1.42)
0.096
Unrelated BMT
1.34 (1.10–1.62)
0.003
UCBT
1.65 (1.38–1.98)
<0.001
Grade III–IV acute GVHD Related BMT
1.00
Related PBSCT
1.81 (1.27–2.58)
0.001
Unrelated BMT
1.34 (0.91–1.96)
0.140
UCBT
1.64 (1.15–2.34)
0.007
Chronic GVHD Related BMT
1.00
Related PBSCT
1.61 (1.38–1.88)
<0.001
Unrelated BMT
1.13 (0.95–1.34)
0.182
UCBT
0.85 (0.71–1.01)
0.070
Extensive chronic GVHD Related BMT
1.00
Related PBSCT
1.67 (1.36–2.05)
<0.001
Unrelated BMT
1.14 (0.91–1.43)
0.254
31
UCBT
0.59 (0.46–0.76)
<0.001
Relapse Related BMT
1.00
Related PBSCT
1.02 (0.85–1.24)
0.798
Unrelated BMT
0.74 (0.60–0.91)
0.005
UCBT
0.61 (0.50–0.75)
<0.001
Non-relapse mortality Related BMT
1.00
Related PBSCT
1.45 (1.13–1.86)
0.003
Unrelated BMT
1.43 (1.11–1.85)
0.006
UCBT
2.24 (1.77–2.84)
<0.001
Relapse-free survival Related BMT
1.00
Related PBSCT
1.19 (1.02–1.38)
0.025
Unrelated BMT
0.96 (0.82–1.13)
0.616
UCBT
1.19 (1.03–1.38)
0.022
Overall survival Related BMT
1.00
Related PBSCT
1.25 (1.07–1.47)
0.006
Unrelated BMT
1.06 (0.89–1.25)
0.519
UCBT
1.41 (1.20–1.65)
<0.001
HR indicates hazard ratio; CI, confidence interval; BMT, bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation; GVHD, graft-versus-host disease. *Presented results are adjusted for patient and transplant characteristics that were significantly associated with each outcome. Relapse-free survival is adjusted for age, sex, performance status and cytogenetic risk; overall survival for age, sex, performance status and cytogenetic risk; relapse for conditioning intensity and cytogenetic risk; non-relapse mortality for age, sex and performance status; neutrophil engraftment for sex mismatch and year of
32
transplantation; platelet engraftment for age, sex and performance status; grade II–IV acute GVHD for conditioning intensity and cytogenetic risk; grade III–IV acute GVHD for age, sex, performance status, conditioning intensity and cytogenetic risk; chronic GVHD for sex, cytogenetic risk, sex mismatch and year of transplantation; and extensive chronic GVHD for sex, cytogenetic risk, sex mismatch and year of transplantation. Full results of the multivariate analyses are provided in Supplementary Table 1.
33
Table 4. Adjusted Effects of Graft Types on Outcomes in a Landmark Analysis at 2 Years after Transplantation Estimated by Multivariate Analysis HR (95% CI)
P-value
Relapse Related BMT
1.00
Related PBSCT
1.52 (0.86–2.71)
0.149
Unrelated BMT
1.07 (0.57–2.00)
0.834
UCBT
0.55 (0.26–1.17)
0.121
Non-relapse mortality Related BMT
1.00
Related PBSCT
1.93 (1.22–3.06)
0.005
Unrelated BMT
0.78 (0.44–1.38)
0.398
UCBT
1.05 (0.62–1.81)
0.847
Relapse-free survival Related BMT
1.00
Related PBSCT
1.79 (1.25–2.58)
0.001
Unrelated BMT
0.93 (0.61–1.41)
0.740
UCBT
0.86 (0.56–1.33)
0.502
Overall survival Related BMT
1.00
Related PBSCT
1.76 (1.20–2.59)
0.004
Unrelated BMT
0.87 (0.56–1.37)
0.552
UCBT
0.91 (0.57–1.44)
0.683
HR indicates hazard ratio; CI, confidence interval; BMT, bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation. *Presented results are adjusted for patient and transplant characteristics that were significantly associated with each outcome. Relapse-free survival is adjusted for conditioning intensity, cytogenetic risk and sex mismatch; overall survival for conditioning intensity, cytogenetic risk and sex mismatch; relapse for performance status and conditioning intensity; and non-relapse mortality for age,
34
cytogenetic risk and sex mismatch. Full results of the multivariate analyses are provided in Supplementary Table 2.
35
Journal Pre-proof
Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation Masamitsu Yanada , Takaaki Konuma , Satoshi Yamasaki , Yachiyo Kuwatsuka , Masayoshi Masuko , Masatsugu Tanaka , Yukiyasu Ozawa , Takashi Toya , Takahiro Fukuda , Shuichi Ota , Masashi Sawa , Naoyuki Uchida , Hirohisa Nakamae , Tetsuya Eto , Junya Kanda , Minoko Takanashi , Yoshinobu Kanda , Yoshiko Atsuta , Shingo Yano PII: DOI: Reference:
S1083-8791(19)30656-1 https://doi.org/10.1016/j.bbmt.2019.09.036 YBBMT 55749
To appear in:
Biology of Blood and Marrow Transplantation
Received date: Accepted date:
2 July 2019 27 September 2019
Please cite this article as: Masamitsu Yanada , Takaaki Konuma , Satoshi Yamasaki , Yachiyo Kuwatsuka , Masayoshi Masuko , Masatsugu Tanaka , Yukiyasu Ozawa , Takashi Toya , Takahiro Fukuda , Shuichi Ota , Masashi Sawa , Naoyuki Uchida , Hirohisa Nakamae , Tetsuya Eto , Junya Kanda , Minoko Takanashi , Yoshinobu Kanda , Yoshiko Atsuta , Shingo Yano , Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation, Biology of Blood and Marrow Transplantation (2019), doi: https://doi.org/10.1016/j.bbmt.2019.09.036
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of the American Society for Transplantation and Cellular Therapy
HIGHLIGHTS
This study aimed to investigate time-varying effects of graft type on allogeneic HCT outcomes for AML.
RFS and OS were worse for related PBSCT and UCBT than for related BMT.
Adverse impact of UCBT was observed only during the early phase of transplantation.
Adverse impact of related PBSCT continued even after two years post-transplant.
Our findings raise concerns regarding the increased risk of late non-relapse mortality with the use of PBSC grafts.
1
Time-Varying Effects of Graft Type on Outcomes for Patients with Acute Myeloid Leukemia Undergoing Allogeneic Hematopoietic Cell Transplantation
Masamitsu Yanada1, Takaaki Konuma2, Satoshi Yamasaki3, Yachiyo Kuwatsuka4, Masayoshi Masuko5, Masatsugu Tanaka6, Yukiyasu Ozawa7, Takashi Toya8, Takahiro Fukuda9, Shuichi Ota10, Masashi Sawa11, Naoyuki Uchida12, Hirohisa Nakamae13, Tetsuya Eto14, Junya Kanda15, Minoko Takanashi16, Yoshinobu Kanda17, Yoshiko Atsuta18,19, Shingo Yano20
1
Aichi Cancer Center, Nagoya, Japan; 2The Institute of Medical Science, The
University of Tokyo, Tokyo, Japan; 3National Hospital Organization Kyusyu Medical Center, Fukuoka, Japan;4Nagoya University Hospital, Nagoya, Japan; 5
Niigata University Medical and Dental Hospital, Niigata, Japan; 6Kanagawa
Cancer Center, Yokohama, Japan; 7Japanese Red Cross Nagoya First Hospital, Nagoya, Japan; 8Tokyo Metropolitan Komagome Hospital, Tokyo, Japan; 9
National Cancer Center Hospital, Tokyo, Japan; 10Sapporo Hokuyu Hospital,
Sapporo, Japan; 11Anjo Kosei Hospital, Anjo, Japan;
12
Toranomon Hospital,
Tokyo, Japan; 13Graduate School of Medicine, Osaka City University, Osaka, Japan; 14Hamanomachi Hospital, Fukuoka, Japan;
15
Graduate School of
Medicine, Kyoto University, Kyoto, Japan; 16Japanese Red Cross Society, Tokyo, Japan; 17Jichi Medical University, Shimotsuke, Japan;
18
for Hematopoietic Cell Transplantation, Nagoya, Japan;
Japanese Data Center 19
Nagoya University
Graduate School of Medicine, Nagoya, Japan; 20The Jikei University School of Medicine, Tokyo, Japan
2
Corresponding author: Masamitsu Yanada, MD, PhD Address: Department of Hematology and Cell Therapy, Aichi Cancer Center, 1-1 Kanokoden, Chikusa-ku, Nagoya, 464-8681, Japan TEL: +81-52-762-6111 FAX: +81-52-764-2967 E-mail: [email protected]
Short title: Time-varying effects of graft type in allo-HCT
Keywords: acute myeloid leukemia, allogeneic hematopoietic cell transplantation, bone marrow transplantation, peripheral blood stem cell transplantation, umbilical cord blood transplantation, late effect
Financial disclosure: This work was supported in part by a grant from the Practical Research Project for Allergic Diseases and Immunology (Research Technology of Medical Transplantation) from the Japan Agency for Medical Research and Development (AMED), grant number: 18ek0510023h0002, and a grant from the Aichi Cancer Research Foundation.
Conflict of interest statement: There are no conflicts of interest to report.
ABSTRACT
3
This study aimed to investigate time-varying effects of graft type on outcomes for patients with acute myeloid leukemia (AML) undergoing allogeneic hematopoietic cell transplantation. For this purpose, we analyzed 3952 patients comprising 720 of whom underwent matched related bone marrow transplantation (BMT), 1004 matched related peripheral blood stem cell transplantation (PBSCT), 856 matched unrelated BMT, and 1372 umbilical cord blood transplantation (UCBT). The 4-year relapse-free survival (RFS) rates were 59.1%, 52.8%, 59.5%, and 50.6%, respectively. Compared to related BMT, related PBSCT, unrelated BMT and UCBT were associated with higher risk of non-relapse mortality, and unrelated BMT and UCBT with lower risk of relapse. As a result, both RFS and overall survival (OS) were comparable between related BMT and unrelated BMT, but were worse for related PBSCT and UCBT than for related BMT. Adverse impact of UCBT was observed only during the early phase of transplantation, whereas that of related PBSCT continued even after two years post-transplant. Our findings raise concerns about the increased risk of late non-relapse mortality with the use of PBSC grafts, and suggest that related BMT is preferable to related PBSCT, and matched unrelated BMT is the next choice in the absence of a matched related donor.
4
INTRODUCTION Allogeneic hematopoietic cell transplantation (HCT) is an effective treatment that makes long-term survival possible for patients with high-risk acute myeloid leukemia (AML) [1-3]. Because of the highly complex nature of the modality, however, success depends on multiple factors. The graft type, i.e., related and unrelated bone marrow (BM), peripheral blood stem cells (PBSC), or umbilical cord blood (UCB), is among the most significant transplantation-related factors, and has been extensively investigated in a number of prospective and retrospective studies [4-23]. The Cox proportional hazards model is widely used in survival analysis to assess the prognostic impact of individual covariates [24]. This method is reliable assuming that proportionality of the hazards does not change with time. However, the effect of graft type on transplantation outcomes may be time-dependent, because the graft type affects a variety of outcome events including engraftment, graft-versus-host disease (GVHD), and relapse due to the graft-versus-leukemia (GVL) effect, each of which occurs at different time points after transplantation. This study aimed to investigate time-varying effects of graft type on outcomes for patients with AML undergoing allogeneic HCT. To this end, we used landmark analysis and additive hazards model analysis, the former for evaluating late effects [24], and the latter for estimating how the effect of covariates changes over time [25].
5
METHODS Study Population All data were collected through the Transplant Registry Unified Management Program sponsored by the Japanese Society for Hematopoietic Cell Transplantation and the Japanese Data Center for Hematopoietic Cell Transplantation. This registration program covers nearly all transplantation centers nationwide, and each participating center is required to report annually follow-up information for consecutively registered HCTs [26-28]. Patients with AML (except acute promyelocytic leukemia) were selected from the database if they were aged 16 years or older, were in first or second complete remission (CR) at the time of transplantation, and had undergone their first allogeneic HCT between 2000 and 2016 from a matched sibling donor, a matched unrelated donor, or a UCB unit that met the criteria specified below. For a matched sibling donor, HLA-A, -B, and -DR at the antigen level were taken into consideration, as were HLA-A, -B, -C and -DRB1 at the allele level for a matched unrelated donor. The UCB unit had to be a single-unit containing a total nucleated cell (TNC) dose 2.0 × 107/kg of the recipient’s body weight and show at least 4/6-match at the antigen level for HLA-A, -B, and -DR, in line with clinical practice in Japan [29]. Patients receiving PBSC from an unrelated donor were not included, because unrelated PBSC transplantation (PBSCT) was rarely performed in Japan during the study period [30]. The study presented here
6
therefore compared four graft types: related BM, related PBSC, unrelated BM, and UCB. This study was approved by the Institutional Review Board of the Aichi Cancer Center, and was conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from each patient.
Definitions Cytogenetic risk was classified in accordance with the criteria specified by the National Comprehensive Cancer Network Guidelines as described in detail elsewhere [31]. The conditioning intensity was determined in line with consensus criteria [32]. Neutrophil and platelet engraftment was defined as the first of three consecutive measures with a neutrophil count 0.5 × 109/L and a platelet count 50 × 109/L. Acute and chronic GVHD were assessed according to the standard criteria [33,34]. For the analysis of chronic GVHD, only patients who had survived 100 days post-transplant without relapse were included.
Statistical Analysis The primary objective of this study was to assess the effect of graft type on relapse-free survival (RFS), and the secondary objective was to assess such effects on overall survival (OS), relapse, and non-relapse mortality. RFS was defined as the time from transplantation to relapse, death or last visit, and OS as the time from transplantation to death or last visit. The probabilities of RFS and
7
OS were estimated by means of the Kaplan–Meier estimator. Multivariate analyses were performed using the Cox proportional hazards model, and hazard ratios (HRs) were calculated together with the corresponding 95% confidence intervals (CIs). The final multivariate model was built using a backward stepwise model selection approach, with a significance cutoff level of 0.05 for removing factors from the model. The variable for graft type, as the main effect, was entered into the final model irrespective of its statistical significance. Relapse and non-relapse mortality were considered to be mutually competing events, and their incidence was estimated with the aid of the cumulative incidence estimator. The Fine and Gray model was used for multivariate analyses. Cumulative incidence of neutrophil and platelet engraftment, and GVHD was also calculated by accommodating competing risks (death for neutrophil and platelet engraftment, and death and relapse for GVHD). Landmark analysis was performed to assess the impact of prognostic variables on RFS, OS, relapse and non-relapse mortality for patients alive without relapse 2 years post transplantation. The additive hazards model proposed by Aalen was used for analysis of time-dependent changes in the effect of graft type on RFS [25]. All statistical analyses were performed with Stata version 14.2 software (StataCorp. LP, College Station, TX, USA).
RESULTS Patients
8
A total of 3952 patients were eligible for analysis, of whom 720 underwent related BM transplantation (BMT), 1004 related PBSCT, 856 unrelated BMT, and 1372 UCB transplantation (UCBT). The median follow-up for surviving patients was 4.4 years (0.1–17.3 years). Table 1 summarizes patient characteristics according to graft type. Several differences were noted among the four groups. Patients undergoing HCT from a related donor were more likely to be transplanted during first CR (78% for related BMT, 79% for related PBSCT, 71% for unrelated BMT, and 69% for UCBT). UCBT recipients were older and received reduced-intensity conditioning more frequently. The median age of patients was 43 years for related BMT, 44 for related PBSCT, 46 for unrelated BMT, and 51 for and UCBT, and reduced-intensity conditioning was used for 19%, 25%, 21%, and 36%, respectively. Transplantation from a female donor to a male recipient was less common for unrelated BMT (26% for related BMT, 30% for related PBSCT, 14% for unrelated BMT, and 29% for UCBT). The percentage of patients transplanted in the early years covered by this study (2000–2010 vs 2011–2016, where the median was used as a cutoff) was 63% for related BMT, 51% for related PBSCT, 58% for unrelated BMT, and 39% for UCBT. For GVHD prophylaxis, anti-thymocyte globulin (ATG) was used in only 1.8% of the overall study population. Owing to the limited number, the use of ATG was not considered for further analyses.
Engraftment and GVHD
9
Table 2 summarizes cumulative incidences of engraftment and GVHD according to graft type, and Table 3 adjusted effects of graft type using related BMT as the reference. In multivariate analysis, the likelihood of neutrophil and platelet engraftment was higher for related PBSCT, while it was lower for UCBT. Unrelated BMT showed a similar likelihood of neutrophil engraftment, but a lower likelihood of platelet engraftment. Related PBSCT was associated with higher risk of grade III–IV acute GVHD as well as overall and extensive chronic GVHD. Unrelated BMT was at higher risk of grade II–IV acute GVHD, whereas the risks of grade III–IV acute GVHD and chronic GVHD were not significantly different from those for related BMT. UCBT was associated with higher risk of grade II–IV and grade III–IV acute GVHD, but the risk of extensive chronic GVHD was lower conversely. Full results of the multivariate analysis are provided in Supplementary Table 1.
Relapse and Non-Relapse Mortality Cumulative incidences of relapse and non-relapse mortality according to graft type are summarized in Table 2 and graphically displayed in Figures 1A and 1B. The cumulative incidence of relapse was similar for related BMT and related PBSCT, but lower for unrelated BMT and UCBT (27.8%, 28.3%, 21.1%, 18.8% at 4 years, respectively; Figure 1A). The cumulative incidence of non-relapse mortality was highest for UCBT and lowest for related BMT, with 12.9% for related BMT, 18.9% for related PBSCT, 19.3% for unrelated BMT, and 30.6% for
10
UCBT (Figure 1B). Multivariate analysis showed that, compared to related BMT, related PBSCT was associated with equivalent risk of relapse and higher risk of non-relapse mortality, unrelated BMT with lower risk of relapse and higher risk of non-relapse mortality, and UCBT with lower risk of relapse and higher risk of non-relapse mortality (Table 3). Full results of the multivariate analysis are provided in Supplementary Table 1.
Relapse-Free and Overall Survival The probabilities of RFS and OS according to graft type are summarized in Table 2 and graphically displayed in Figures 1C and 1D. The RFS curves were identical for related BMT and unrelated BMT, but lower for related PBSCT and UCBT (Figure 1C). The 4-year RFS rates were 59.1% for related BMT, 52.8% for related PBSCT, 59.5% for unrelated BMT, and 50.6% for UCBT. Similar trends were observed for OS (Figure 1D), with corresponding rates of 64.5%, 58.2%, 62.7%, and 53.9% at 4 years. Multivariate analysis using related BMT as the reference showed that both RFS and OS were inferior for related PBSCT and UCBT, and were comparable for unrelated BMT (Table 3). Full results of the multivariate analysis are provided in Supplementary Table 1. Effects of several covariates on RFS, the primary endpoint of this study, were assessed within individual graft type groups. RFS was not affected by year of transplantation (1992–2010 or 2011–2016) for each group (P = 0.180 for related BMT, P = 0.714 for related PBSCT, P = 0.752 for unrelated BMT, and P = 0.666 for UCBT).
11
As for unrelated BMT, although the donor age was significantly higher by 4 years in the later cohort (P < 0.001), there was no difference in RFS between patients whose donor age was below or above the median value of 36 years (P = 0.140). For patients undergoing UCBT, RFS did not differ by whether they received a UCB unit containing a TNC dose below or above the median value of 2.67 × 107/kg of the recipient’s body weight (P = 0.199).
Outcomes after a 2-Year Landmark Period Landmark analysis was performed for assessment of the late effects of graft type on outcomes. The landmark was set at 2 years after transplantation, and the 1946 patients who were alive without relapse at this time point were analyzed. The median follow-up for surviving patients was 3.5 years (0.0–15.3 years) after the landmark. The cumulative incidence of relapse was 3.2% for related BMT, 7.5% for related PBSCT, 4.0% for unrelated BMT, and 3.9% for UCBT at 3 years from the 2-year landmark (Figure 2A). There was no difference in the risk of relapse between related BMT and other transplantations in multivariate analysis (Table 4). The cumulative incidence of non-relapse mortality was 7.1% for related BMT, 10.7% for related PBSCT, 3.2% for unrelated BMT, and 7.7% for UCBT (Figure 2B) at 3 years from the 2-year landmark. Multivariate analysis showed that only related PBSCT significantly correlated with higher risk of non-relapse mortality compared to related BMT (Table 4). The RFS and OS rates at 3 years after the 2-year landmark were
12
89.7% and 90.7% for related BMT, 81.8% and 85.4% for related PBSCT, 92.8% and 94.3% for unrelated BMT, and 88.5% and 90.9% for UCBT, respectively (Figures 2C and 2D). Multivariate analysis showed that only related PBSCT was associated with worse RFS and OS than related BMT, while unrelated BMT and UCBT had comparable RFS and OS with related BMT. Full results of the multivariate analysis are provided in Supplementary Table 2.
Changes in the Effect of Graft Type on RFS We next used the additive hazards model to examine how the effect of graft type on RFS changes over time. Figure 3 shows the cumulative coefficients of each graft type versus related BMT. The plots display changes in hazard rates over time, with an ascending slope indicating increased hazard rates, a descending slope reduced hazard rates, and a flat slope no time-varying effect. The results have been adjusted for factors associated with RFS, i.e., age, sex, performance status and cytogenetic risk. The curve for related PBSCT fluctuated for the first 1 year, and continued to rise thereafter (Figure 3A), indicating that the risk for related PBSCT increased steadily after 1 year. The curve for unrelated BMT increased for the first half year, then decreased, and had a roughly flat slope until 4 years, after which it decreased further (Figure 3B). This indicates that the risk for unrelated BMT increased in the early period of around 6 months post-transplantation, but decreased after that. The curve for UCBT increased steeply for the first few months and then increased gradually for the
13
first one and a half years. The slope then changed from positive to negative, and became flat thereafter (Figure 3C). This suggests that the risk for UCBT increased substantially in the early period, especially for the first few months, but there was no further increase in the risk in the late period.
DISCUSSION The study presented here show that graft type significantly affected engraftment, GVHD, relapse and non-relapse mortality, which collectively impacted survival outcomes in a time-dependent manner. Adverse impacts of unrelated BMT and UCBT were observed only during the early phase of transplantation, whereas that of related PBSCT continued even after two years post-transplant. PBSC has now largely replaced BM as stem cell source for allogeneic HCT. Although its use for unrelated HCT is still very limited in Japan because unrelated PBSCT has been used in clinical practice since 2010 [30], the annual number is expected to increase. Owing to the inclusion of more CD34+ cells and T-cells in the graft, PBSCT is characterized by faster engraftment but also higher risk of GVHD compared to BMT. While GVHD is a major complication associated with allogeneic HCT, it is accompanied by a potentially protective effect against relapse, and a few randomized studies have shown a reduction in relapse with PBSCT versus BMT [4,17]. Our analysis, which focused on patients with AML in CR, shows that the use of related PBSCT increased non-relapse
14
mortality without a reduction in relapse, leading to worse RFS and OS compared to related BMT. Furthermore, the adverse impact of related PBSCT on non-relapse mortality, RFS and OS remained even after 2 years post-transplant, with the cumulative hazard curve for RFS ascending steadily thereafter. Related PBSCT was also associated with significantly higher risk of chronic GVHD compared to related BMT. These results taken together suggest that higher risk of chronic GVHD for related PBSCT resulted in later mortality but did not help prevent relapse for our patients. In addition, related PBSCT was associated with increased risk of grade III–IV acute GVHD. Despite the lack of data on unrelated PBSCT for this study, it is conceivable that higher risk of GVHD is a matter of particular concern when choosing unrelated PBSCT instead of unrelated BMT as previously reported [18]. Since only approximately 40% of transplantation candidates have a matched related donor [19,35,36], an alternative donor needs to be found for the remaining 60% to undergo allogeneic HCT. With the aid of high-resolution HLA typing, there have been significant improvements in the outcome of unrelated HCT [37,38]. Several studies have shown that allogeneic HCT from an unrelated donor matched at the allele level yields results similar to those of allogeneic HCT from a matched related donor [10,13-15], and our findings are consistent with those previously reported. A higher risk of non-relapse mortality for our patients undergoing unrelated BMT was counterbalanced by reduced risk of relapse, resulting in RFS and OS similar to those for related BMT. Although the risk of
15
grade II–IV acute GVHD was higher for unrelated BMT, there was no significant increase in the risk of grade III–IV acute GVHD or chronic GVHD. The cumulative hazard curve on RFS peaked around the first half year, and no late effect was observed. These results support the notion that a matched unrelated donor is a reasonable alternative both in the short- and long-term when a matched related donor is not available. Here, attention should be paid to the fact that the longer time needed for unrelated donor search sometimes precludes performing a transplantation during CR. On this point, our previous decision analysis showed that unrelated BMT is preferable to immediate UCBT, if unrelated BMT can be performed within 3–9 months after achieving first CR [20]. UCB is another stem cell source for allogeneic HCT. In addition to its immediate availability, UCB has the benefit of greater tolerance of HLA disparity. On the other hand, the low number of hematopoietic stem cells contained in a UCB unit constitutes a major limitation to its use, which can lead to delayed engraftment, higher risk of post-transplant infection, and early death. Our data show that UCBT had a more potent protective effect against relapse than related BMT, but this advantage was outweighed by higher risk of non-relapse mortality, leading to significantly worse RFS and OS. While at the same time, the risk of extensive chronic GVHD was lower, and there was no late adverse effect on survival outcomes. The cumulative hazard curve for RFS peaked at one and half years despite a steep increase soon after transplantation. These findings indicate that the adverse impact of UCBT is limited mostly to the early phase of
16
transplantation. One may argue that our criterion for UCB unit selection of TNC dose 2.0 × 107/kg was less stringent than those used in Western countries, which allowed for an inclusion of patients who received TNC doses that were considered inadequate. In this regard, however, we had previously shown that the TNC dose in excess of the minimum requirement of 2.0 × 107/kg were not predictive for OS, relapse, or non-relapse mortality [29]. Moreover, our finding has been recently corroborated by a joint study by the European Society for Blood and Marrow Transplantation and the Japanese Society of Hematopoietic Cell Transplantation, in which the TNC dose showed no significant effect on OS [39]. At any rate, our data highlight that the reduction in non-relapse mortality during the early phase of transplantation is the key to the success of UCBT. Several investigators have implemented strategies, including double-unit UCBT [40] and ex vivo expansion of cord blood cells [41]. These efforts, in concert with the establishment of better unit selection criteria as well as improvements in supportive care measures, are expected to overcome the current limitations of UCBT. Selection bias must be considered when interpreting our results. Which patients received which grafts was the result of non-randomized selection, and we do not know, for example, whether patients who underwent UCBT did so as planned initially or only after the failure of an unrelated donor search. It is also possible that our results were confounded by changes in practice implemented during the study period, including patient selection, expansion of the unrelated
17
donor registry, the accumulated experience of each transplantation center, and advances in transplantation procedures and supportive care measures. Although the large number of patients assessed in our study allowed for accurate multivariate analysis by adjusting for the confounding effects of known factors, such unknown or unmeasured factors may have affected the results. This type of bias can theoretically be resolved by a prospective randomized study; however, conducting such a study is unfeasible. This situation enables retrospective studies such as ours to serve as important sources of information. While acknowledging potential limitations inherent to a registry-based study, an inclusion of a homogeneous patient population, i.e., adult patients with AML in first or second CR, represents the unique strength of this study. Additional caution needs to be exercised in that our findings were based exclusively on Japanese patient data. Because racial differences in the risk of GVHD and thus possibly of GVL effect are well recognized [42,43], it may be premature to apply our conclusions to other patient populations. Finally, although our registry data did not contain information on costs or the quality of life (QOL), it is conceivable that the graft type may affect these outcomes. Late complications such as chronic GVHD have been shown to deteriorate patients’ QOL [44,45], and increase medical costs [46,47]. Addressing how the graft type impacts these outcomes will provide additional important insights into graft selection. In conclusion, this study presents the time-varying effects of graft type on outcomes after allogeneic HCT. Our results raise a particular concern about the
18
increased risk of late non-relapse mortality with the use of PBSC grafts, and suggest that related BMT is preferable to related PBSCT, and BMT from a matched unrelated donor is the next choice for patients with AML in CR who need allogeneic HCT but lack a matched related donor. Future studies that deal with allogeneic HCT using various types of grafts should give careful consideration to the proportional hazards assumption.
REFERENCES 1.
Gupta V, Tallman MS, Weisdorf DJ. Allogeneic hematopoietic cell transplantation for adults with acute myeloid leukemia: myths, controversies, and unknowns. Blood. 2011;117:2307-2318.
2.
Yanada M. Allogeneic hematopoietic cell transplantation for acute myeloid leukemia during first complete remission: a clinical perspective. Int J Hematol. 2015;101:243-254.
3.
Cornelissen JJ, Blaise D. Hematopoietic stem cell transplantation for patients with AML in first complete remission. Blood. 2016;127:62-70.
4.
Powles R, Mehta J, Kulkarni S, et al. Allogeneic blood and bone-marrow stem-cell transplantation in haematological malignant diseases: a randomised trial. Lancet. 2000;355:1231-1237.
5.
Vigorito AC, Marques Junior JF, Aranha FJ, Oliveira GB, Miranda EC, De Souza CA. A randomized, prospective comparison of allogeneic bone marrow and peripheral blood progenitor cell transplantation in the treatment of hematologic malignancies: an update. Haematologica. 2001;86:665-666.
6.
Couban S, Simpson DR, Barnett MJ, et al. A randomized multicenter comparison of bone marrow and peripheral blood in recipients of matched sibling allogeneic transplants for myeloid malignancies. Blood.
19
2002;100:1525-1531. 7.
Laughlin MJ, Eapen M, Rubinstein P, et al. Outcomes after transplantation of cord blood or bone marrow from unrelated donors in adults with leukemia. N Engl J Med. 2004;351:2265-2275.
8.
Rocha V, Labopin M, Sanz G, et al. Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. N Engl J Med. 2004;351:2276-2285.
9.
Moore J, Nivison-Smith I, Goh K, et al. Equivalent survival for sibling and unrelated donor allogeneic stem cell transplantation for acute myelogenous leukemia. Biol Blood Marrow Transplant. 2007;13:601-607.
10.
Schetelig J, Bornhauser M, Schmid C, et al. Matched unrelated or matched sibling donors result in comparable survival after allogeneic stem-cell transplantation in elderly patients with acute myeloid leukemia: a report from the cooperative German Transplant Study Group. J Clin Oncol. 2008;26:5183-5191.
11.
Atsuta Y, Suzuki R, Nagamura-Inoue T, et al. Disease-specific analyses of unrelated cord blood transplantation compared with unrelated bone marrow transplantation in adult patients with acute leukemia. Blood. 2009;113:1631-1638.
12.
Nagafuji K, Matsuo K, Teshima T, et al. Peripheral blood stem cell versus bone marrow transplantation from HLA-identical sibling donors in patients with leukemia: a propensity score-based comparison from the Japan Society for Hematopoietic Stem Cell Transplantation registry. Int J Hematol. 2010;91:855-864.
13.
Walter RB, Pagel JM, Gooley TA, et al. Comparison of matched unrelated and matched related donor myeloablative hematopoietic cell transplantation for adults with acute myeloid leukemia in first remission. Leukemia. 2010;24:1276-1282.
14.
Gupta V, Tallman MS, He W, et al. Comparable survival after HLA-well-matched unrelated or matched sibling donor transplantation for acute myeloid leukemia in first remission with unfavorable cytogenetics at diagnosis. Blood. 2010;116:1839-1848.
20
15.
Schlenk RF, Dohner K, Mack S, et al. Prospective evaluation of allogeneic hematopoietic stem-cell transplantation from matched related and matched unrelated donors in younger adults with high-risk acute myeloid leukemia: German-Austrian trial AMLHD98A. J Clin Oncol. 2010;28:4642-4648.
16.
Friedrichs B, Tichelli A, Bacigalupo A, et al. Long-term outcome and late effects in patients transplanted with mobilised blood or bone marrow: a randomised trial. Lancet Oncol. 2010;11:331-338.
17.
Mielcarek M, Storer B, Martin PJ, et al. Long-term outcomes after transplantation of HLA-identical related G-CSF-mobilized peripheral blood mononuclear cells versus bone marrow. Blood. 2012;119:2675-2678.
18.
Anasetti C, Logan BR, Lee SJ, et al. Peripheral-blood stem cells versus bone marrow from unrelated donors. N Engl J Med. 2012;367:1487-1496.
19.
Yanada M, Kurosawa S, Yamaguchi T, et al. Effect of related donor availability on outcome of AML in the context of related and unrelated hematopoietic cell transplantation. Bone Marrow Transplant. 2013;48:390-395.
20.
Yanada M, Kanda J, Ohtake S, et al. Unrelated bone marrow transplantation or immediate umbilical cord blood transplantation for patients with acute myeloid leukemia in first complete remission. Eur J Haematol. 2016;97:278-287.
21.
Terakura S, Atsuta Y, Tsukada N, et al. Comparison of outcomes of 8/8 and 7/8 allele-matched unrelated bone marrow transplantation and single-unit cord blood transplantation in adults with acute leukemia. Biol Blood Marrow Transplant. 2016;22:330-338.
22.
Savani BN, Labopin M, Blaise D, et al. Peripheral blood stem cell graft compared to bone marrow after reduced intensity conditioning regimens for acute leukemia: a report from the ALWP of the EBMT. Haematologica. 2016;101:256-262.
23.
Versluis J, Labopin M, Ruggeri A, et al. Alternative donors for allogeneic hematopoietic stem cell transplantation in poor-risk AML in CR1. Blood
21
Adv. 2017;1:477-485. 24.
Delgado J, Pereira A, Villamor N, Lopez-Guillermo A, Rozman C. Survival analysis in hematologic malignancies: recommendations for clinicians. Haematologica. 2014;99:1410-1420.
25.
Aalen OO. A linear regression model for the analysis of life times. Stat Med. 1989;8:907-925.
26.
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.
27.
Atsuta Y. Introduction of Transplant Registry Unified Management Program 2 (TRUMP2): scripts for TRUMP data analyses, part I (variables other than HLA-related data). Int J Hematol. 2016;103:3-10.
28.
Kanda J. Scripts for TRUMP data analyses. Part II (HLA-related data): statistical analyses specific for hematopoietic stem cell transplantation. Int J Hematol. 2016;103:11-19.
29.
Yanada M, Konuma T, Kuwatsuka Y, et al. Unit selection for umbilical cord blood transplantation for adults with acute myeloid leukemia in complete remission: a Japanese experience. Bone Marrow Transplant. 2019 May 8. [Epub ahead of print].
30.
Saito H, Ito M, Kato S, et al. The Japan Marrow Donor Program, 25 years of experience in achieving 20,000 bone marrow transplantations: organization structure, activity, and financial basis. Bone Marrow Transplant. 2018;53:609-616.
31.
Yanada M, Mori J, Aoki J, et al. Effect of cytogenetic risk status on outcomes for patients with acute myeloid leukemia undergoing various types of allogeneic hematopoietic cell transplantation: an analysis of 7812 patients. Leuk Lymphoma. 2018;59:601-609.
32.
Giralt S, Logan B, Rizzo D, et al. Reduced-intensity conditioning for unrelated donor progenitor cell transplantation: long-term follow-up of the first 285 reported to the national marrow donor program. Biol Blood Marrow Transplant. 2007;13:844-852.
33.
Sullivan KM, Agura E, Anasetti C, et al. Chronic graft-versus-host disease
22
and other late complications of bone marrow transplantation. Semin Hematol. 1991;28:250-259. 34.
Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15:825-828.
35.
Mawad R, Gooley TA, Sandhu V, et al. Frequency of allogeneic hematopoietic cell transplantation among patients with high- or intermediate-risk acute myeloid leukemia in first complete remission. J Clin Oncol. 2013;31:3883-3888.
36.
Labopin M, Gorin NC, Polge E, et al. A prospective registration study to determine feasibility of hematopoietic SCT in adults with acute leukemia: planning, expectations and reality. Bone Marrow Transplant. 2014;49:376-381.
37.
Majhail NS, Chitphakdithai P, Logan B, et al. Significant improvement in survival after unrelated donor hematopoietic cell transplantation in the recent era. Biol Blood Marrow Transplant. 2015;21:142-150.
38.
Yanada M, Masuko M, Mori J, et al. Patients with acute myeloid leukemia undergoing allogeneic hematopoietic cell transplantation: trends in survival during the past two decades. Bone Marrow Transplant. 2019;54:578-586.
39.
Kanda J, Hayashi H, Ruggeri A, et al. Prognostic factors for adult single cord blood transplantation among European and Japanese populations: the Eurocord/ALWP-EBMT and JSHCT/JDCHCT collaborative study. Leukemia. 2019.
40.
Barker JN, Weisdorf DJ, DeFor TE, et al. Transplantation of 2 partially HLA-matched umbilical cord blood units to enhance engraftment in adults with hematologic malignancy. Blood. 2005;105:1343-1347.
41.
de Lima M, McNiece I, Robinson SN, et al. Cord-blood engraftment with ex vivo mesenchymal-cell coculture. N Engl J Med. 2012;367:2305-2315.
42.
Oh H, Loberiza FR, Jr., Zhang MJ, et al. Comparison of graft-versus-host-disease and survival after HLA-identical sibling bone marrow transplantation in ethnic populations. Blood. 2005;105:1408-1416.
23
43.
Morishima Y, Kawase T, Malkki M, et al. Significance of ethnicity in the risk of acute graft-versus-host disease and leukemia relapse after unrelated donor hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2013;19:1197-1203.
44.
Pidala J, Kurland B, Chai X, et al. Patient-reported quality of life is associated with severity of chronic graft-versus-host disease as measured by NIH criteria: report on baseline data from the Chronic GVHD Consortium. Blood. 2011;117:4651-4657.
45.
Kurosawa S, Oshima K, Yamaguchi T, et al. Quality of Life after Allogeneic Hematopoietic Cell Transplantation According to Affected Organ and Severity of Chronic Graft-versus-Host Disease. Biol Blood Marrow Transplant. 2017;23:1749-1758.
46.
Saito AM, Cutler C, Zahrieh D, et al. Costs of allogeneic hematopoietic cell transplantation with high-dose regimens. Biol Blood Marrow Transplant. 2008;14:197-207.
47.
Khera N, Emmert A, Storer BE, Sandmaier BM, Alyea EP, Lee SJ. Costs of allogeneic hematopoietic cell transplantation using reduced intensity conditioning regimens. Oncologist. 2014;19:639-644.
24
FIGURE LEGENDS
Figure 1. Cumulative incidence of relapse (A), cumulative incidence of non-relapse mortality (B), relapse-free survival (C), and overall survival (D) by graft type. Patients undergoing matched related bone marrow transplantation (BMT; N = 720), matched related peripheral blood stem cell transplantation (PBSCT; N = 1,004), matched unrelated BMT (N = 856), and umbilical cord blood transplantation (UCBT; N = 1,372) are compared.
25
Figure 2. Cumulative incidence of relapse (A), cumulative incidence of non-relapse mortality (B), relapse-free survival (C), and overall survival (D) by graft type in a landmark analysis at 2 years after transplantation. Only those who survived without relapse for at least 2 years are included. Patients undergoing matched related bone marrow transplantation (BMT; N = 384), matched related peripheral blood stem cell transplantation (PBSCT; N = 510), matched unrelated BMT (N = 482), and umbilical cord blood transplantation (UCBT; N = 570) are compared.
26
27
Figure 3. Cumulative coefficients in terms of relapse-free survival (RFS) obtained from the additive hazards model for related peripheral blood stem cell transplantation (PBSCT) (A), unrelated bone marrow transplantation (BMT) (B), and umbilical cord blood transplantation (UCBT) (C), in comparison to related BMT. The plots display changes in hazard rates over time, with an ascending slope indicating increased hazard rates, a descending slope reduced hazard rates, and a flat slope no time-varying effect. Results are adjusted for age, sex, performance status and cytogenetic risk, all of which were identified as factors significantly associated with RFS.
28
Table 1. Characteristics of Patients According to Graft Type Related BMT
Related PBSCT
Unrelated BMT
UCBT
N = 720
N = 1004
N = 856
N = 1372
Age Median (range), years
43
(16–73)
44
(16–73)
46
(16–70)
51
(16–85)
<50 years
483
(67%)
632
(63%)
505
(59%)
633
(46%)
50 years
237
(33%)
372
(37%)
351
(41%)
739
(54%)
Male
396
(55%)
596
(59%)
482
(56%)
705
(51%)
Female
324
(45%)
408
(41%)
374
(44%)
666
(49%)
0–1
695
(97%)
976
(98%)
784
(98%)
1251
(96%)
2
20
(3%)
23
(2%)
19
(2%)
58
(4%)
Myeloablative
543
(81%)
719
(75%)
671
(79%)
876
(64%)
Reduced-intensity
124
(19%)
237
(25%)
175
(21%)
496
(36%)
Favorable
120
(17%)
155
(15%)
136
(16%)
162
(12%)
Intermediate
444
(62%)
645
(64%)
545
(64%)
891
(65%)
Poor
102
(14%)
128
(13%)
101
(12%)
233
(17%)
54
(8%)
76
(8%)
74
(9%)
86
(6%)
CR1
558
(78%)
793
(79%)
610
(71%)
942
(69%)
CR2
162
(23%)
211
(21%)
246
(29%)
430
(31%)
Female to male
186
(26%)
294
(30%)
121
(14%)
327
(29%)
Others
532
(74%)
700
(70%)
732
(86%)
805
(71%)
2000–2010
458
(63%)
511
(51%)
495
(58%)
538
(39%)
2011–2016
262
(36%)
493
(49%)
361
(42%)
834
(69%)
Sex
ECOG performance status
Conditioning
Cytogenetic risk
Unevaluable Disease status
Sex mismatch
Year of transplantation
BMT indicates bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation; ECOG, Eastern Cooperative Oncology Group; CR, complete remission.
29
Table 2. Post-transplant Outcomes by Graft Type
Related
Related
Unrelated
BMT
PBSCT
BMT
95.5%
98.4%
95.9%
78.3%
89.2%
91.5%
86.5%
56.7%
25.6%
28.9%
33.2%
36.5%
6.2%
10.2%
9.2%
10.5%
Chronic GVHD, at 2 years
34.6%
51.5%
38.8%
26.6%
Extensive chronic GVHD,
19.7%
33.0%
23.2%
11.5%
Relapse, at 4 years
27.8%
28.3%
21.1%
18.8%
Non-relapse mortality, at 4
12.9%
18.9%
19.3%
30.6%
59.1%
52.8%
59.5%
50.6%
64.5%
58.2%
62.7%
53.9%
Neutrophil engraftment, at
UCBT
30 days Platelet engraftment, at 60 days Grade II–IV acute GVHD, at 100 days Grade III–IV acute GVHD, at 100 days
at 2 years
years Relapse-free survival, at 4 years Overall survival, at 4 years
BMT indicates bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation; GVHD, graft-versus-host disease.
30
Table 3. Adjusted Effects of Graft Types on Outcomes Estimated by Multivariate Analysis HR (95% CI)
P-value
Neutrophil engraftment Related BMT
1.00
Related PBSCT
1.89 (1.68–2.12)
<0.001
Unrelated BMT
0.95 (0.86–1.05)
0.327
UCBT
0.43 (0.40–0.48)
<0.001
Platelet engraftment Related BMT
1.00
Related PBSCT
1.52 (1.34–1.73)
<0.001
Unrelated BMT
0.82 (0.73–0.92)
0.001
UCBT
0.33 (0.30–0.37)
<0.001
Grade II–IV acute GVHD Related BMT
1.00
Related PBSCT
1.17 (0.97–1.42)
0.096
Unrelated BMT
1.34 (1.10–1.62)
0.003
UCBT
1.65 (1.38–1.98)
<0.001
Grade III–IV acute GVHD Related BMT
1.00
Related PBSCT
1.81 (1.27–2.58)
0.001
Unrelated BMT
1.34 (0.91–1.96)
0.140
UCBT
1.64 (1.15–2.34)
0.007
Chronic GVHD Related BMT
1.00
Related PBSCT
1.61 (1.38–1.88)
<0.001
Unrelated BMT
1.13 (0.95–1.34)
0.182
UCBT
0.85 (0.71–1.01)
0.070
Extensive chronic GVHD Related BMT
1.00
Related PBSCT
1.67 (1.36–2.05)
<0.001
Unrelated BMT
1.14 (0.91–1.43)
0.254
31
UCBT
0.59 (0.46–0.76)
<0.001
Relapse Related BMT
1.00
Related PBSCT
1.02 (0.85–1.24)
0.798
Unrelated BMT
0.74 (0.60–0.91)
0.005
UCBT
0.61 (0.50–0.75)
<0.001
Non-relapse mortality Related BMT
1.00
Related PBSCT
1.45 (1.13–1.86)
0.003
Unrelated BMT
1.43 (1.11–1.85)
0.006
UCBT
2.24 (1.77–2.84)
<0.001
Relapse-free survival Related BMT
1.00
Related PBSCT
1.19 (1.02–1.38)
0.025
Unrelated BMT
0.96 (0.82–1.13)
0.616
UCBT
1.19 (1.03–1.38)
0.022
Overall survival Related BMT
1.00
Related PBSCT
1.25 (1.07–1.47)
0.006
Unrelated BMT
1.06 (0.89–1.25)
0.519
UCBT
1.41 (1.20–1.65)
<0.001
HR indicates hazard ratio; CI, confidence interval; BMT, bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation; GVHD, graft-versus-host disease. *Presented results are adjusted for patient and transplant characteristics that were significantly associated with each outcome. Relapse-free survival is adjusted for age, sex, performance status and cytogenetic risk; overall survival for age, sex, performance status and cytogenetic risk; relapse for conditioning intensity and cytogenetic risk; non-relapse mortality for age, sex and performance status; neutrophil engraftment for sex mismatch and year of
32
transplantation; platelet engraftment for age, sex and performance status; grade II–IV acute GVHD for conditioning intensity and cytogenetic risk; grade III–IV acute GVHD for age, sex, performance status, conditioning intensity and cytogenetic risk; chronic GVHD for sex, cytogenetic risk, sex mismatch and year of transplantation; and extensive chronic GVHD for sex, cytogenetic risk, sex mismatch and year of transplantation. Full results of the multivariate analyses are provided in Supplementary Table 1.
33
Table 4. Adjusted Effects of Graft Types on Outcomes in a Landmark Analysis at 2 Years after Transplantation Estimated by Multivariate Analysis HR (95% CI)
P-value
Relapse Related BMT
1.00
Related PBSCT
1.52 (0.86–2.71)
0.149
Unrelated BMT
1.07 (0.57–2.00)
0.834
UCBT
0.55 (0.26–1.17)
0.121
Non-relapse mortality Related BMT
1.00
Related PBSCT
1.93 (1.22–3.06)
0.005
Unrelated BMT
0.78 (0.44–1.38)
0.398
UCBT
1.05 (0.62–1.81)
0.847
Relapse-free survival Related BMT
1.00
Related PBSCT
1.79 (1.25–2.58)
0.001
Unrelated BMT
0.93 (0.61–1.41)
0.740
UCBT
0.86 (0.56–1.33)
0.502
Overall survival Related BMT
1.00
Related PBSCT
1.76 (1.20–2.59)
0.004
Unrelated BMT
0.87 (0.56–1.37)
0.552
UCBT
0.91 (0.57–1.44)
0.683
HR indicates hazard ratio; CI, confidence interval; BMT, bone marrow transplantation; PBSCT, peripheral blood stem cell transplantation; UCBT, umbilical cord blood transplantation. *Presented results are adjusted for patient and transplant characteristics that were significantly associated with each outcome. Relapse-free survival is adjusted for conditioning intensity, cytogenetic risk and sex mismatch; overall survival for conditioning intensity, cytogenetic risk and sex mismatch; relapse for performance status and conditioning intensity; and non-relapse mortality for age,
34
cytogenetic risk and sex mismatch. Full results of the multivariate analyses are provided in Supplementary Table 2.
35