|Home | About | Journals | Submit | Contact Us | Français|
The risk of acute graft-versus-host disease (GVHD) after HLA-matched sibling bone marrow (BM) transplantation is lower in Japanese than in Caucasian patients. However, race may have differential effect on GVHD dependent on the graft source. North American Caucasian and Japanese patients receiving their first allogeneic BM or peripheral blood stem cell (PBSC) transplantations from an HLA-matched sibling for leukemia were eligible. BM was used in 13% and 53% of Caucasian and Japanese patients, respectively. In multivariate analysis, the interaction term between race and graft source was not significant in any of the models, indicating that graft source does not affect the impact of race on outcomes. The risk of grades III–IV acute GVHD was significantly lower in Japanese than in Caucasian patients (hazard ratio (HR) 0.74, 95% confidence interval (CI) 0.57–0.96), which resulted in lower risk of non-relapse mortality in Japanese patients (HR 0.69, 95% CI 0.54–0.89). The risk of relapse was also lower in this group. Lower risk of non-relapse mortality and relapse resulted in lower overall mortality rates among Japanese patients. In conclusion, irrespective of graft source, the risk of severe acute GVHD is lower in Japanese patients, which results in lower risk of non-relapse mortality.
Several studies have demonstrated lower acute graft-versus-host disease (GVHD) risks after human leukocyte antigen (HLA)-matched sibling bone marrow transplantation (BMT) among Japanese, a genetically homogeneous population, compared to Caucasian and other races.[1,2] The lower incidence of acute GVHD was likely due to the relative homogeneity of minor histocompatibility antigens among Japanese.[3,4] Understanding outcome differences among races is particularly important for interpretation of transplant study findings and application of these results to clinical practice. In recent years, despite evidence of slightly increased risk of GVHD, peripheral blood stem cells (PBSC) have been more frequently used than bone marrow (BM) as stem cell grafts because of earlier neutrophil and platelet engraftment, perceived better overall survival in advanced or high-risk diseases, and easier collection procedure. The effect of race may differ according to stem cell sources, because the number and subpopulations of T and other immune cells responsible for GVHD differ significantly between BM and PBSC.[6,7] To our knowledge, no other reports have assessed outcome differences between races after HLA-matched sibling PBSC transplantation (PBSCT).
Several groups in Europe and the U.S. have conducted randomized controlled trials comparing HLA-matched sibling BMT and PBSCT.[8–11] Based on these randomized studies, the Stem Cell Trialists’ Collaborative Group conducted an individual-patient data meta-analysis using combined data from 1,111 adult patients. The incidence of grades III–IV acute GVHD was 26% and 21% in the HLA-matched sibling PBSCT and BMT groups, respectively, with an odds ratio of 1.39 (95% confidence interval [CI], 1.03–1.88). Regarding matched unrelated donor transplantation, Anasetti et al. recently compared outcomes of unrelated donor BMT and PBSCT in a large multicenter randomized trial. They reported a similar incidence of acute GVHD among BMT and PBSCT recipients. Nagafuji et al. retrospectively evaluated outcomes after HLA-matched sibling BMT and PBSCT in Japan using Japanese national registry data. In this study, the incidence of grades III–IV acute GVHD was 14% and 5% in the HLA-matched sibling PBSCT and BMT, respectively, with a hazards ratio of 2.23 (95% CI, 1.04–4.78). These reports suggest that the incidence of severe acute GVHD was very low after HLA-matched sibling BMT in the Japanese population and the risk ratio of severe acute GVHD in PBSCT vs. BMT was much larger in Japanese than in Caucasian and other populations (2.23 vs. 1.39). In this context, we hypothesized that the lower incidence of acute GVHD in Japanese compared to Caucasian populations observed in HLA-matched sibling BMT will not be detectable in HLA-matched sibling PBSCT, and that there is an interaction between race and use of BM or PBSC. Furthermore, we hypothesized that this difference in severe acute GVHD rates between the races will affect other transplant outcomes and choice of BM or PBSC in HLA-matched sibling transplantation. Therefore, we accessed the large registry data of the Center for International Blood and Marrow Transplant Research (CIBMTR) and the Japan Society of Hematopoietic Cell Transplantation (JSHCT) to compare transplant outcomes between Japanese and Caucasian patients who received BMT or PBSCT from an HLA-matched sibling.
Data were obtained from the CIBMTR and the Transplant Registry Unified Management Program (TRUMP) of JSHCT for U.S. and Japanese transplants, respectively. The CIBMTR is a combined research program of the Medical College of Wisconsin and the National Marrow Donor Program. The CIBMTR comprises a voluntary network of more than 500 transplantation centers worldwide, which contribute detailed data on consecutive allogeneic and autologous hematopoietic cell transplantation to a centralized statistical center. Observational studies conducted by the CIBMTR are performed in compliance with all applicable U.S. federal regulations pertaining to the protection of human research participants. Protected health information used in the performance of such research is collected and maintained in the CIBMTR capacity as a public health authority under the Health Insurance Portability and Accountability Act Privacy Rule. Additional details regarding the data source are described elsewhere. JSHCT developed the TRUMP to enable transplant centers to manage patient information with emphases on convenience to centers, safety of patient information, and quality of data management. TRUMP unifies data of 4 Japanese registries; JSHCT, Japanese Society of Pediatric Hematology, Japan Marrow Donor Program, and Japan Cord Blood Bank Network. Adult related allogeneic transplantations performed in Japan are registered through JSHCT. Registration of the JSHCT does not require patient consent because clinical information is anonymized. The study was approved by the Institutional Review Board of the National Marrow Donor Program, Medical College of Wisconsin, Saitama Medical Center, Jichi Medical University, and the Data Management Committees of the JSHCT.
North American Caucasian and Japanese patients who received their first allogeneic BMT or PBSCT from an HLA-matched sibling for acute myelogenous leukemia (AML) and acute lymphoblastic leukemia (ALL) in first and second complete remission or chronic myelogenous leukemia (CML) in complete remission or chronic or accelerated phase between 2000 and 2011 were eligible. Patients were 18–59 years of age, received a myeloablative conditioning regimen, and used cyclosporine or tacrolimus-based GVHD prophylaxis. Patients with use of in vivo or ex vivo T cell depletion methods were excluded. AML or ALL in first complete remission and CML in first chronic phase or complete remission were defined as early risk disease. AML or ALL in the second complete remission and CML in the second or higher chronic phase or accelerated phase were defined as intermediate risk disease. Eastern Cooperative Oncology Group performance statuses of 0 and 1 or more were converted to Karnofsky performance status (KPS) 90–100 and <90, respectively, for patients in the Japanese dataset lacking KPS scores, following a conversion formula used in the CIBMTR. In the CIBMTR registration forms, race is categorized into “White or Caucasian”, “Black or African American”, “American Indian or Alaska Native”, “Asian”, and “Native Hawaiian or other Pacific Islander”. North American Caucasian includes Caucasians who live in the U.S., Canada, and Mexico.
Outcomes assessed include acute GVHD, chronic GVHD, relapse, non-relapse mortality, leukemia-free survival, and overall survival. Acute and chronic GVHDs were diagnosed and graded using the traditional criteria.[16,17] Non-relapse mortality was defined as death without relapse. Leukemia-free survival was defined as being alive in remission. For overall survival, death from any cause was the event.
Patient-, disease-, and transplant-related factors were compared between groups classified by race. Cohorts were compared using Chi-square and Kruskal-Wallis tests for categorical variables and continuous variables, respectively. Cumulative incidences for GVHD, relapse, and non-relapse mortality were calculated using the cumulative incidence function to account for competing risks. Competing events were death without relapse for relapse, relapse for non-relapse mortality, and death without GVHD for acute and chronic GVHD. Gray’s test was employed to evaluate the overall differences among cumulative incidence functions. Leukemia-free and overall survival rates were estimated using the Kaplan-Meier estimator. The log-rank test was used to compare survival experience among different groups. Ninety-five percentage CIs for true survival probabilities were calculated based on arcsine-square root transformation. The impact of race and stem cell source on outcomes of interest was evaluated using the Cox proportional hazards model. The assumption of proportional hazards for each covariate in the Cox model was tested using time-dependent covariates. When the test indicated differential effects over time (non-proportional hazards), models were constructed to break the post-transplant time course into two periods. The maximized partial likelihood method was used to determine the most appropriate breakpoint. The proportionality assumptions were then further tested. A stepwise selection procedure was used to identify covariates associated with outcomes. The main effects of this study - race (North American Caucasian vs. Japanese) and graft source (BM vs. PBSC) - were included in all steps of model building. Other variables considered included recipient age group, recipient sex, KPS, disease status prior to the transplantation, donor age, donor-recipient cytomegalovirus (CMV) status, donor-recipient sex match, donor-recipient ABO blood group match, conditioning regimen, GVHD prophylaxis, and year of transplantation. Covariates that reached a significance level of 0.05 were included in the final models. An interaction test was used to determine whether the effects of graft source on transplant outcomes were different between the two groups. If the interaction was statistically significant, the main effects were combined into one main effect variable (graft source by race) in the final models. Otherwise, graft source and race were included as two independent main effects in the final models. Any potential interactions between main effect term and variables in the final model were evaluated. SAS 9.3 (SAS Inc., Cary, NC) statistical software was used for all analyses.
Table 1 shows patient, donor, and transplant characteristics. BM was used in 13% and 53% of the Caucasian and Japanese patients, respectively. Japanese patients were more likely to be young, have better KPS scores, have early stages of ALL, have both donor and recipient CMV seropositivity, receive transplants from ABO mismatched donors, receive a cyclophosphamide + TBI regimen, receive cyclosporine-based prophylaxis, and receive a transplant after 2006. Majority of North American Caucasian is American and only 3% is Canadian.
An interaction test was used to determine whether the effects of graft source on transplant outcomes were different between the two populations. In multivariate analysis, the interaction term between race and graft source was not significant in any of the models for all outcomes analyzed, which indicated that the impact of race on outcomes did not differ according to graft source (Supplemental Table 1). Therefore, graft source and race were treated as two independent main effects analyzing all outcomes.
The unadjusted cumulative incidence of grades II–IV acute GVHD at 100 days post transplantation was 31% (95% CI, 29%–34%) and 31% (95% CI, 28%–33%) in Caucasian and Japanese patients, respectively (Gray’s test, P=0.11); the incidence for grades III–IV acute GVHD post transplantation was 14% (95% CI, 12%–16%) and 8% (95% CI, 7%–10%) in Caucasian and Japanese patients, respectively (Gray’s test, P<0.001). In multivariate analysis, the risk of grades II–IV acute GVHD was comparable between both groups (hazard ratio [HR] 1.04, 95% CI 0.87–1.24, P=0.677), whereas the risk of grades III–IV acute GVHD was significantly lower in Japanese patients compared to Caucasians (HR 0.75, 95% CI 0.57–0.99, P=0.042) (Table 2). The risk of chronic GVHD was significantly lower in Japanese compared to Caucasian patients (within 12 months after transplantation, HR 0.85, 95% CI 0.73–0.99, P = 0.037; 12 months or more after transplantation, HR 0.33, 95% CI 0.21–0.51, P < 0.001). The unadjusted cumulative incidences of GVHD stratified by stem cell sources are shown in Supplemental Table 2. Adjusted cumulative incidence curves are shown in Figure 1.
The unadjusted cumulative incidence of relapse 3 years post transplantation was 33% (95% CI, 30%–35%) and 25% (95% CI, 23%–28%) in Caucasian and Japanese patients, respectively (Gray’s test, P<0.001); the incidence of non-relapse mortality post transplantation was 18% (95% CI, 16%–20%) and 13% (95% CI, 11%–15%), respectively (Gray’s test, P<0.001). In multivariate analysis, the risks of relapse and non-relapse mortality were significantly lower in Japanese than in Caucasians patients (relapse; HR 0.73, 95% CI 0.61–0.87, P<0.001, non-relapse mortality; HR 0.72, 95% CI 0.55–0.93, P=0.012, Table 3). The adjusted cumulative incidence curves are shown in Figure 2.
The unadjusted probability of disease-free survival 3 years post transplantation was 49% (95% CI, 46%–52%) and 62% (95% CI, 59%–65%) in Caucasian and Japanese patients, respectively (log-rank test, P<0.001), and 57% (95% CI, 54%–60%) and 68% (95% CI, 65%–71%) for overall survival post transplantation, respectively (log-rank test, P<0.001). In multivariate analysis, the risks of death or relapse and overall mortality were significantly lower in Japanese than Caucasian patients (death or relapse; HR 0.78, 95% CI 0.68–0.89, P<0.001, overall mortality; HR 0.72, 95% CI 0.60–0.85, P<0.001, Table 3). Adjusted survival curves are shown in Figure 3.
Contrary to our initial hypothesis, the impact of race on acute GVHD did not differ according to BM or PBSC graft. Irrespective of graft source, the risk of severe acute GVHD was lower in Japanese patients. Although the main determinant of GVHD occurrence is the major histocompatibility antigen, occurrence of GVHD in transplantation from an HLA-identical sibling indicates that non-HLA gene polymorphisms play an important role in GVHD. Candidate gene studies have shown that various gene polymorphisms in recipients and donors are associated with GVHD. Furthermore, the genome-wide association study confirms that some gene polymorphisms such as IL-2, IL-6, CTLA4, HPSE, and MTHFR are responsible for GVHD. Such polymorphisms could be responsible for racial differences in GVHD occurrence. The relative homogeneity of minor histocompatibility antigens among Japanese individuals may also explain the reduced GVHD incidence in this population. Differences in frequencies of HLA antigens or haplotypes between the two populations may also contribute to differences in GVHD incidence. Specific haplotypes have been associated with risk of acute GVHD. The strength of T cell activation caused by minor histocompatibility antigens may differ by haplotype and major histocompatibility complex. Notably, our study showed that differences between the two populations were not affected by use of PBSC, which contains larger T cell populations and smaller non-hematopoietic populations than BM.
Unmeasured factors such as socioeconomic, environmental, or practice differences may also be responsible for outcome differences after allogeneic transplantation in the two populations. Analysis of unrelated myeloablative transplantation using the CIBMTR data indicated that lower income is associated with higher rates of non-relapse and overall mortality, highlighting the importance of socioeconomic factors on transplant outcomes. Health insurance systems also differ; health insurance is privately purchased insurance or social welfare insurance in the U.S., while Japan offers universal health insurance coverage. This difference could affect access to medical care and medicine availability, including recently approved or unapproved treatments. The difficulty of adjusting for these factors limits the interpretation of our findings. Comparison of transplant outcomes of Japanese populations living outside Japan to those living in Japan could help evaluate these effects, although such an analysis is unlikely to be feasible due to the small sample size. Missing or insufficient data is also needs to be considered in interpreting the result. Missing data in CMV serostatus in the Japanese populations is not negligible and may have influenced the findings.
The lower risk of non-relapse mortality in the Japanese patients compared to the Caucasian patients partly reflects the lower incidence of grades III–IV acute GVHD among the Japanese patients. Interestingly, even lower risk of relapse was also observed in the Japanese. This lower risk was previously noted in unrelated BMT and PBSCT. Racial differences in leukemic cell susceptibility to chemotherapeutic agents may play a role in relapse. Yang et al. showed that genetic characteristics associated with Native American ancestry affect the risk of ALL relapse. Furthermore, a particular SNP associated with drug resistance in vitro is also highly associated with relapse. Genetic differences related to relapse between the two populations is worthy of further evaluation in future studies. Another possibility is differences in graft-versus-leukemia effects. Unlike previous findings in cohorts receiving transplantations between 1990 and 1999, we did not observe any difference in the risk of grades II–IV acute GVHD between the two cohorts receiving transplantations between 2000 and 2011. This may be partly because Japanese physicians tended to weaken or taper GVHD prophylaxis earlier than before, expecting to enhance graft-versus-leukemia effects, since the GVHD incidence was very low in an HLA-matched sibling BMT among Japanese patients. Alternatively, GVHD prophylaxis might have been intensified in the decade for Caucasian patients. Although cyclosporine has still been mainly used as GVHD prophylaxis for matched sibling transplantation in Japanese patients, tacrolimus-based GVHD prophylaxis has replaced this in Caucasian patients. Although the type of GVHD prophylaxis did not affect acute GVHD in the present study, such a change of practice may decrease the incidence of acute GVHD and the associated graft-versus-leukemia effects in Caucasian patients. The graft-versus-leukemia effects need to be interpreted according to each disease. In the present study, there was no interaction between race or stem cell source and disease in the multivariate analysis for relapse, which suggested that the impact of race on relapse was independent of disease type. Finally, although we included only early and intermediate disease status, i.e. AML and ALL in 1st or 2nd complete remission and CML in complete remission, or chronic or accelerated phase, lack of information on the cytogenetic and molecular risks in each disease may limit the interpretation of this study.
Contrary to the BMT and PBSCT findings, there was no difference in acute GVHD incidence or overall mortality after pediatric unrelated cord blood transplantation between Japanese and U.S. Caucasian patients. In addition to qualitative and quantitative differences in the composition of BM, PBSC, and unrelated cord blood grafts, this may be partly because advantages associated with shared minor histocompatibility antigens in the more homogenous Japanese population may be weakened in unrelated cord blood transplantation with multiple major histocompatibility antigen mismatches.
In conclusion, irrespective of graft source, the risk of severe acute GVHD in this study was lower in Japanese patients, which resulted in a lower risk of non-relapse mortality. The difference in the incidence of grades III–IV acute GVHD may have been due to racial differences in non-HLA gene polymorphisms and minor histocompatibility antigens presented by specific major histocompatibility antigens. However, this could also be affected by socioeconomic or environmental factors or differences between practices. Although this study cannot separately evaluate genetic and non-genetic effects, it highlights the importance of considering racial differences not only in HLA-matched sibling BMT but also in PBSCT, particularly when interpreting and applying findings from different racial populations.
The impact of race on transplant outcomes was independent of graft source.
The risk of severe acute GVHD was lower in Japanese than in Caucasian patients.
Lower risk of acute GVHD resulted in lower risk of non-relapse mortality in Japanese.
The risk of relapse was lower in Japanese than in Caucasian patients.
Overall survival was higher in Japanese than in Caucasian patients.
CIBMTR Support List
The CIBMTR is supported by Public Health Service Grant/Cooperative Agreement 5U24-CA076518 from the National Cancer Institute (NCI), the National Heart, Lung and Blood Institute (NHLBI) and the National Institute of Allergy and Infectious Diseases (NIAID); a Grant/Cooperative Agreement 5U10HL069294 from NHLBI and NCI; a contract HHSH250201200016C with Health Resources and Services Administration (HRSA/DHHS); two Grants N00014-13-1-0039 and N00014-14-1-0028 from the Office of Naval Research; and grants from Alexion; *Amgen, Inc.; Anonymous donation to the Medical College of Wisconsin; Be the Match Foundation; *Bristol Myers Squibb Oncology; *Celgene Corporation; *Chimerix, Inc.; Fred Hutchinson Cancer Research Center; Gamida Cell Ltd.; Genentech, Inc.; Genzyme Corporation; *Gilead Sciences, Inc.; Health Research, Inc. Roswell Park Cancer Institute; HistoGenetics, Inc.; Incyte Corporation; *Jazz Pharmaceuticals, Inc.; Jeff Gordon Children’s Foundation; The Leukemia & Lymphoma Society; The Medical College of Wisconsin; Merck & Co, Inc.; Mesoblast; *Millennium: The Takeda Oncology Co.; *Miltenyi Biotec, Inc.; National Marrow Donor Program; Neovii Biotech NA, Inc.; Novartis Pharmaceuticals Corporation; Onyx Pharmaceuticals; Optum Healthcare Solutions, Inc.; Otsuka America Pharmaceutical, Inc.; Otsuka Pharmaceutical Co, Ltd. – Japan; Oxford Immunotec; Perkin Elmer, Inc.; Pharmacyclics; *Sanofi US; Seattle Genetics; Sigma-Tau Pharmaceuticals; *Spectrum Pharmaceuticals, Inc.; St. Baldrick’s Foundation; *Sunesis Pharmaceuticals, Inc.; Swedish Orphan Biovitrum, Inc.; Telomere Diagnostics, Inc.; TerumoBCT; Therakos, Inc.; University of Minnesota; and *Wellpoint, Inc. The views expressed in this article do not reflect the official policy or position of the National Institute of Health, the Department of the Navy, the Department of Defense, Health Resources and Services Administration (HRSA) or any other agency of the U.S. Government.
This work was supported in part by the Practical Research Project for Allergic Diseases and Immunology (Research Technology of Medical Transplantation) from Japan Agency for Medical Research and Development, AMED.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. 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.
Disclosure of Conflicts of Interest:
There is no conflict to disclose regarding this paper.
Junya Kanda, Ruta Brazauskas, Zhen-Huan Hu, Yachiyo Kuwatsuka, Yoshiko Atsuta, Wael Saber designed the study, analyzed data and wrote the manuscript.
Koji Nagafugi, Heiwa Kanamori, Yoshinobu Kanda, Koichi Miyamura, Makoto Murata, Takahiro Fukuda, Hisashi Sakamaki, Fumihiko Kimura, Sachiko Seo, Mahmoud Aljurf, Ayami Yoshimi, Giuseppe Milone, William A. Wood, Celalettin Ustun, Shahrukh Hashimi, Marcelo Pasquini, Carmem Bonfim, Jignesh Dalal, and Theresa Hahn designed the study and wrote the manuscript.