Advances in hematopoietic stem cell transplantation in the Asia-Pacific region: the second report from APBMT 2005–2015

Between 2005 and 2015, 138,165 hematopoietic stem cell transplantation (HSCT) were reported in 18 countries/regions in the Asia-Pacific region. In this report, we describe current trends in HSCT throughout the Asia-Pacific region and differences among nations in this region and various global registries. Since 2008, more than 10,000 HSCTs have been recorded each year by the Asia-Pacific Blood and Marrow Transplantation Group Data Center. Between 2005 and 2015, the greatest increase in the number of HSCTs was observed in Vietnam. Allogeneic HSCT was performed more frequently than autologous HSCT, and a majority of cases involved related donors. Regarding allogeneic HSCT, the use of cord blood has remained steady, especially in Japan, and the number of cases involving related HLA non-identical donors has increased rapidly, particularly in China. The incidence of hemoglobinopathy, a main indication for allogeneic HSCT in India, China, Iran, and Pakistan, increased nearly six-fold over the last decade. Among the 18 participating countries/regions, the transplant rate per population varied widely according to the absolute number of HSCTs and the national/regional population size. We believe that this report will not only benefit the AP region but will also provide information about HSCT to other regions worldwide.


Introduction
Hematopoietic stem cell transplantation (HSCT) is the curative treatment modality of choice for many malignant and non-malignant hematologic disorders, and the annual global frequency of this procedure has increased steadily over the past decade [1]. In 2010, Yoshimi et al. published the first report of the number of HSCTs performed between 1986 and 2006 in nine countries/regions in the Asia-Pacific (AP) region [2]. Since then, the Asia-Pacific Blood and Marrow Transplantation Group (APBMT) Data Center has continuously collated the HSCT cases reported in each participating country. The APBMT Activity Survey, which comprises items regarding the HSCT type, donor type, stem cell source, and disease type, is used for collation. Subsequently, the survey data are used to analyze the latest trends in HSCT throughout the AP region, promote HSCT in both emerging economies and advanced countries, and collaborate on various international research studies.
In the last decade, the options for donor selection and the disease indications for HSCT have been expanded [1]. This report aims to summarize the APBMT data and compare the findings to those from Western countries with the intent to identify trends in HSCT in the AP region.

Materials and methods
This detailed retrospective analysis was based on APBMT Activity Survey Data collected from 2005 to 2015. We set 2005 as the earliest time point because some newly participating countries/regions in the APBMT Activity Survey (Australia and New Zealand) had already reported activities prior to the publication of our first report [2]. We also used partial data accumulated since 1986 to summarize activity in this region. At the end of December 2017, 18 of 21 APBMT member countries/regions had submitted data from 2005 to 2015 to the APBMT Data Center.
Six countries/regions (Australia, New Zealand, India, Republic of Korea [referred to as Korea in this paper], Japan, and Taiwan) have their own national registries: the Australasian Bone Marrow Transplant Recipient Registry (ABMTRR), Indian Stem Cell Transplant Registry (ISCTR), Korean Society of Blood and Marrow Transplantation (KSBMT), Japan Society for Hematopoietic Cell Transplantation (JSHCT)/Japanese Data Center for Hematopoietic Cell Transplantation (JDCHCT), and Taiwan Society of Blood and Marrow Transplantation (TBMT), respectively. These countries/regions submitted national registry data every year. In China, Iran, Malaysia, and the Philippines, a single contact person from a major transplant center collated the number of entire HSCTs performed in their country/region each year and submitted those data (subsequently, national registry systems were created in China and Malaysia). Hospitals or centers in Bangladesh, Hong Kong, Myanmar, Pakistan, Singapore, Sri Lanka, Thailand, and Vietnam sent data individually to the APBMT Data Center, which aggregated these data by country. By the end of 2017, the APBMT Data Center had not received data from Cambodia, Indonesia, and Mongolia, although certain HSCT activities were known to have occurred in Indonesia and Mongolia.
The number of HSCTs was collated according to HSCT type, donor type, stem cell source, and disease type. If one patient underwent two transplants in 1 year, the APBMT counted this case as two HSCTs. HSCTs involving multiple stem cell sources during a single procedure were conventionally counted as one HSCT and categorized as a multiple stem cell source transplant; these cases included bone marrow (BM) plus peripheral blood stem cells (PB), BM plus cord blood cells (CB), PB plus CB, or BM plus PB plus CB. Although the APBMT Data Center had not previously used the word "haploidentical", we considered the term "HLA non-identical family" to be roughly synonymous to "haploidentical." Among the disease indications, this survey classified thalassemia, sickle cell disease, and other hemoglobinopathies as hemoglobinopathies.
To compare trends in HSCT over time, the rate of increase was calculated by dividing the frequency of HSCT during 2015 by the frequency in 2005. The transplant rates in each country/region were calculated as the number of each type of HSCT per 10 million residents in 2015. The total population for each country was extracted from a United States Census Bureau report [3], and the gross domestic product (GDP) and GDP per capita were extracted from World Bank data [4]. All analyses in this report were conducted at the APBMT Data Center. The present survey was approved by the institutional review board of the Aichi Medical University School of Medicine.

Overview of trends in HSCT and center numbers
Eighteen of 21 APBMT member countries/regions reported their annual HSCT activities to the APBMT Data Center every year between 2005 and 2015 (Table 1). Centers in Bangladesh, Myanmar, and Sri Lanka began performing HSCTs in 2014 after their initial participation in the Worldwide Network for Blood and Marrow Transplantation/World Health Organization (WBMT/WHO) Workshop in Hanoi, 2011 [5]. The total number of centers in the AP region in 2015 was 624, and the numbers of centers varied by country/region, ranging from one center each in Bangladesh, Myanmar, and Sri Lanka to 373 centers in Japan.
The annual number of HSCTs performed in the AP region has increased continuously each year. The annual number has exceeded 10,000 each year since 2008, and nearly 200,000 total HSCTs were performed from 1986 [2] to 2015. A comparison of changes in the rates of HSCTs from 2005 to 2015 revealed the greatest increase in Vietnam (13.0), followed by China (8.8), the Philippines (7.0), India (6.2), Malaysia (2.7), Taiwan (2.6), and Pakistan (2.5) ( Table 1).

Allogeneic vs. autologous transplantation
Allogeneic HSCT was performed more frequently than autologous HSCT, and the gap between these transplant types increased each year (Fig. 1a). The proportion of allogeneic HSCTs ranged widely among the 18 registered countries/regions, from 32.7% in New Zealand to 86.7% in Pakistan. Only four countries/regions recorded more cases of autologous HSCT vs allogeneic HSCT in 2015 (Australia, Malaysia, New Zealand, and Thailand). Bangladesh and Myanmar performed no allogeneic HSCTs in 2015, which was attributed to the initiation of HSCT after participation in the 1st WBMT/WHO Workshop in 2011 and observation of the recommendations made by the WBMT [6] (Fig. 1b).

Donor types and stem cell sources
As shown in Fig. 2, which comprises data compiled during the first report [2], the number of HSCTs with related donors was consistently higher than the number of HSCTs with unrelated donors. Although the gap in the frequencies of these two donor types decreased from 2004 to 2011, it  S in g a p o r e S r i L a n k a T a iw a n T h a il a n d V ie t n a m Australia, Japan, and New Zealand were the only countries that performed more HSCTs involving unrelated donors ( Table 2). An analysis of the main donor types in allogeneic transplantation cases revealed a recent and rapid increase in the number of HSCTs from HLA non-identical related donors, compared to the mild increases in the numbers of HSCTs from identical siblings and unrelated donors. An analysis of the stem cell sources for allogeneic HSCT demonstrated that the number of peripheral blood stem cell transplants (PBSCT) first overtook that of BM transplants (BMT) in 2007. Subsequently, the frequency of PBSCT has increased continuously in the AP region. Furthermore, the number of HSCTs with multiple stem cell sources has increased over the past years, and 95.0% of these cases were performed in China in 2015 ( Table 2). The most popular multiple stem cell source combination was BM plus PB from a human leukocyte antigen (HLA) non-identical family member (60.4%). Furthermore, that among cases with unrelated donors, the number of cord blood transplants (CBT) was equivalent to the total number of cases involving BM and PB in Japan (Table 2).
Notably, 82.6% of all CBTs were performed in Japan. The high frequency of PBSCT in cases of non-malignant diseases was also a characteristic trend observed in the Asia-Pacific region (Table 3).

Indications
The HSCT numbers increased steadily from 2005 to 2015 for all diseases except chronic myelogenous leukemia (CML) (Fig. 3a, b). In 2015, acute myelogenous leukemia (AML) accounted for 24.3% of all HSCTs, followed by lymphoid malignancies (19.6%), plasma cell disorders (18.1%), and acute lymphoid leukemia (ALL) (13.0%). Although hemoglobinopathy accounted for only 2.9% of all HSCTs, the rate of increase in this disease was prominent, compared to those of other major diseases (Fig. 3b). During the past decade, the number of HSCTs performed for hemoglobinopathy increased by more than six-fold, and these procedures were performed in India, China, Iran, Pakistan, Korea, and Malaysia. The numbers of HSCTs performed for CML decreased in 2015 relative to early 2000s in all countries, except India and Malaysia.

Transplant rates
As shown in Table 4, the rates of each type of transplant per population varied widely among the AP countries/regions. Figure 4 depicts the relationships between the absolute number of transplants and GDP/GDP per capita. As shown, the transplant numbers correlated with GDP rather than GDP per capita in the AP region.

Discussion
After Thomas et al. reported comprehensive studies that aimed to obtain successful HSCT outcomes [7,8], some HSCTs were performed in AP region even in the 1970s [9,10]. However, HSCT was not applied regularly until the 1980s [11][12][13][14][15]. In 2007, the APBMT Data Center began to survey HSCT activity in related to procedures performed since 1986. Since the initial APBMT report of HSCT activities published in 2010 [2], the number of countries/ regions reporting their annual data has doubled (from 9 to 18), and the number of HSCT centers increased by 43.8% (from 432 to 621). In 2015, 17,572 HSCTs were performed, representing a 2.3-fold increase in the number of HSCTs performed (from 7771 to 17,572) since 2005. Among the 18 participating countries/regions, the increases in the numbers of reported HSCTs were particularly noticeable in Vietnam, China, the Philippines, and India. These increases were attributed to (1) the establishment and organization of academic activities and international collaborations through the annual APBMT Congresses and WBMT workshops [5] in Vietnam [16] and the Philippines [17] after the late 2000s, and (2) the very large national populations, recent  remarkable economic growth, and advances in medical services in China and India [3,18].
Our results demonstrate the diversity of HSCT activities throughout the AP region. Three countries have just implemented HSCT programs (Bangladesh, Myanmar, and Sri Lanka), while three others have not yet reported HSCT activities (Cambodia, Indonesia, and Mongolia). By contrast, Japan, Korea and Australia have consistently reported more than 1000 HSCTs performed annually since 2005. The number of HSCT centers is extremely high in Japan because HSCT is routinely performed by hematology departments, mostly due to high coverage of the national health insurance scheme. The fluctuations in the number of HSCTs performed in Thailand might result from insufficient communication among the domestic centers in Thailand as well as our data center. Despite the huge disparities in socioeconomic statuses among countries in the AP region [4], HSCT has increased steadily in all member countries of the APBMT in the past decade. Nevertheless, these huge disparities mean that the data in this report may not necessarily reflect all regional trends, but rather the trends of the few countries participating in the survey. Gratwohl et al. reported an association of the economic gap between lowincome and high-income countries/regions with global transplant rates and suggested that increased regional support would foster the growth of HSCT programs [19]. To overcome the potential effects of this disparity, we have specified the names of the corresponding countries/regions where necessary.
An overview of the observed trends reveals a remarkable difference between the data from Western countries and those from the AP region. Namely, allogeneic HSCTs were performed more frequently than autologous HSCT in the AP region. Gratwohl et al. reported that on average allogeneic HSCTs accounted for 38 and 46% of all HSCTs performed in Europe and the Americas, respectively, during 2006-2012 [1,20]. In the AP region, however, 12 of 16 participating countries/regions reported that allogeneic HSCT was performed more frequently than autologous HSCT (Table 2), with allogeneic HSCTs accounting for 63.5% of all procedures in 2015 (excluding data from Bangladesh and Myanmar, which had not yet initiated allogeneic transplantation in 2015). We note that the HSCT rates per population in India and China remained relatively low, despite the relatively large absolute numbers of HSCTs in these countries. Although we might attribute this discrepancy to the large populations of these countries, the absolute numbers of procedures correlated strongly with the GDP of each country.
Our study revealed three characteristic findings regarding donor types and stem cell sources. First, we elucidated the number of PBSCTs from unrelated donors and CBTs in Japan. Although the Japan Marrow Donor Program (JMDP) facilitated large numbers of unrelated HSCTs using BM cells and the number of PBSCTs from related donors was also large, the number of PBSCTs from unrelated donors was markedly small in comparison ( Table 2). We attribute this finding to the specific JMDP facilitation of G-CSFmobilized peripheral blood harvesting from volunteer donors after a 10-years project confirming the safety of family donors [21]. Accordingly, Japan launched an unrelated PBSCT program in 2010. The cumulative number of PBSCTs from unrelated donors remained small at the end of 2015 due to stepwise increases in the performance of this procedure [22]. By contrast, the number of CBTs performed in Japan was large, consistent with the cumulative annual numbers of HSCTs with related or unrelated donors, and increased continuously. Uchida et al. reported that nearly a third of all CBTs worldwide were performed in Japan [23,24], contrary to a recent global decreasing trend in the performance of CBT and concomitant increase in HSCTs from haploidentical donors [25,26]. This discrepancy may be attributable to the relatively small body size of an adult patient; in most cases, therefore, a single unit of CB stem cells can provide sufficient cell numbers for engraftment. Currently, the six Japan Cord Blood Banks aim to collect more than 10 9 /units of CB stem cells, which would cover the majority of adult recipients.
Second, China reported a large number of HSCTs from related donors that involved the co-infusion of stem cells from multiple sources ( Table 2). Based on the long-term experiences with haplo-HSCT according to the Beijing Protocol, Lu, Huang, and colleagues proposed that graftversus-host disease (GVHD) and rejection after haplo-HSCT could be prevented by the co-infusion of G-CSFmobilized BM and PB and the administration of antithymocyte globulin [27,28]. In this region, the relatively large numbers of HSCTs from HLA non-identical family donors are mostly attributable to Chinese cases and may reflect the difficulty of identifying an HLA-matched sibling donor due to the recently recalled single-child policy. Accordingly, China simultaneously developed a HSCT bank system for BM, PB, and CB and a haploidentical transplant technique that differs from the current Posttransplant cyclophosphamide (PT/CY) used in Western countries [29].
Third, the number of cases using PB as a stem cell source overtook those using BM, even for non-malignant diseases (Table 3). This trend is largely attributable to India and differs from the patterns observed in the US and Europe [30]. In India, PBSCT is preferentially selected to improve engraftment in patients with non-malignant diseases under non-radiation regimens (from a personal communication with A Srivastava). We await the outcomes of this trend in India.
The last decade saw increases in the number of HSCTs performed for each disease except for CML. Furthermore, each hematological malignancy was reported as a top indication for HSCT in all AP countries/regions, consistent with reports from other regions [26,31]. Nevertheless, the reported incidences of plasma cell disorders and lymphoid malignancies, the major targets of autologous transplantation, were lower in the AP region than in Western countries [26]. Most APBMT member countries may place a greater priority curative allogeneic transplantation for leukemia and MDS than on prolonging the survival of patients with plasma cell disorders or lymphoid malignancies. Similar trends have been observed for non-malignant hematological diseases. In this context, hemoglobinopathy was among the most important indications for allogeneic HSCT, and the rate of increase in the number of transplants for this disease was remarkably higher than that of other diseases (Fig. 3b). In summary, the definitive indication of allogeneic HSCT for hemoglobinopathy and primary indication of this procedure for leukemia/MDS might partly explain the dominance of allogeneic HSCT in this region.
In conclusion, we observed a wide range of HSCT activities in the AP region. Notably, wide variations were observed in the transplant and donor types, stem cell sources, and disease indications among the member countries/regions. Notably, both the number of AP countries/ regions with centers performing HSCT and the number of HSCTs have increased in the past decade. We believe that the data in this report will not only be useful to AP countries/regions but will also inform other regions worldwide [20]. report was supported by the APBMT Registry Committee and Data Center. We also thank Yukari Nakao and Hiroe Namizaki of the APBMT Data Center for their data management. The secretarial assistance of Haruko Nakamura is also appreciated.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix
All contributing centers that were registered with the APBMT Data Center are listed here (some centers may no longer perform HSCT, may have merged with other centers, or changed their names). Australia