Progress in survival following three decades of allogeneic hematopoietic cell transplantation for myelodysplastic syndrome: A real‐world registry study in Japan

Allogeneic hematopoietic cell transplantation (HCT) can eradicate myelodysplastic syndrome (MDS) clones through high-dose chemoradiotherapy and the graft-versus-tumor effect,1 and it remains the only potentially curative treatment modality for MDS. Although the improvement of transplant procedures, such as advanced supportive care and increased availability of unrelated donors, has resulted in progressive outcomes over time after allogeneic HCT,2-4 the results for trends of transplant outcomes have been unclear for adult patients with MDS. Here, we investigated the trends in outcomes after allogeneic HCT for unselected adult MDS patients in Japan. Details for methods are provided in the supplemental methods (Data S1). We identified 6257 adult patients with MDS who received their first allogeneic HCT between 1987 and 2020. Among them, 550, 1954, and 3753 patients were transplanted in 1987–2000, 2001–2010, and 2011–2020, respectively. Over the three time periods, there was a progressive increase in age at HCT, male recipients, poor karyotype, high-risk at HCT, diagnosis of HCT <6 months, unrelated donor, HLA mismatch, and graft-versus-host disease (GVHD) prophylaxis with methotrexate (Table S1). The absolute annual numbers of HCT for age ≥60 years increased substantially over time, from 1 (1.9%) in 1997 to 245 (51.0%) in 2020 (Figure 1A). The absolute annual number of unrelated bone marrow transplantation (UBMT) has increased since 2004. More recently, the increase in unrelated cord blood transplantation (UCBT) was remarkable, from 1 (1.2%) in 1999 to 158 (32.9%) in 2020 (Figure 1B). As for conditioning intensity, both myeloablative conditioning (MAC) and reduced-intensity conditioning have increased over time, but MAC has been more commonly performed in all periods (Figure 1C). The unadjusted probability of overall survival (OS) was not significant across the time periods (Figure 1D), whereas the unadjusted probability of disease-free survival (DFS) had significantly deteriorated between the time periods (Figure 1F). However, multivariate analyses showed that OS and DFS were significantly improved across the time periods (Table S2). The adjusted probabilities of OS and DFS improved over time (Figure 1E,G). The unadjusted cumulative incidence of relapse deteriorated between the time periods (Figure 1H). In the multivariate analysis, the relapse rate was not significantly associated with the time periods. The adjusted cumulative incidence of relapse was not improved across the time periods (Figure 1I). The unadjusted cumulative incidence of non-relapse mortality (NRM) had significantly improved between the time periods (Figure 1J). Multivariate analyses showed that NRM was significantly improved across the time periods (Table S2). The adjusted cumulative incidence of NRM improved over time (Figure 1K). The most common cause of death was an infection in the first two time periods, but it was relapse or progression of MDS in the most recent period (Table S3). Thus, the proportion of deaths from relapse or progression of MDS increased in a stepwise fashion. Although the proportion of deaths from GVHD, pulmonary complication, organ failure, and hemorrhage decreased in the latter two time periods, the proportion of death from infection was not different across the three time periods. Subgroup analyses showed that OS and DFS had significantly improved over time based on age, disease risk at HCT, donor type, and conditioning intensity, except that OS and DFS did not improve among patients aged 40–59 years and those with low risk at HCT (Figure S1, Table S4). An important finding in our study was that the risk of relapse or progression of MDS after allogeneic HCT has gradually increased over time, and the proportion of death from relapse or progression of MDS has increased in a stepwise fashion. Such an increment of relapse or progression could have arisen because of the increased proportions over time of patients with advanced diseases, such as high-risk at HCT and poor karyotype. Therefore, to reduce the disease burden before allogeneic HCT, induction chemotherapy and hypomethylating agents have been tried in some patients with a higher disease burden. Indeed, the main change in clinical practice during the study period has been the introduction of azacytidine since 2011 in Japan. However, consistent with previous studies,5, 6 our study showed that the relapse rates and mortality did not improve in high-risk patients who received pre-transplant induction chemotherapy and hypomethylating agents compared to those who did not receive them, regardless of transplant year (Figure S2). This may reflect the selection bias that derives from the inclusion of patients with advanced diseases who received pre-transplant induction chemotherapy and hypomethylating agents. However, these pre-transplant treatments could be useful as a bridging therapy to allogeneic HCT that aims to avoid disease progression among patients with high disease burden. Furthermore, post-transplant maintenance strategies using several targeting agents might be beneficial, but their effects remain elusive for the further reduction of relapse or progression of MDS after allogeneic HCT.7 Importantly, a reduction of relapse rate will be needed for the further improvement of allogeneic HCT for MDS. The improvement in survival over time primarily resulted from decreased NRM among MDS patients. This result could be encouraging despite increased indications for older patients and more frequent use of unrelated donors over time. The proportions of death from GVHD, pulmonary complications, organ failure, and hemorrhage decreased in the latter two time periods, but those from infection did not decrease across the three time periods. These findings indicate that the prevention and management of GVHD, pulmonary complications, organ failure, and hemorrhage might be adequate, which is consistent with previous studies,2, 3 but the management of carryover infection, partly due to prolonged neutropenia during the pretransplant phase, might remain insufficient after allogeneic HCT for MDS. Therefore, stricter prevention and treatment of infectious complications could be required for further improvement of NRM after allogeneic HCT for MDS patients. Recently, the BMT CTN 1102 study has prospectively confirmed a survival benefit of allogeneic HCT over non-HCT therapy in high-risk MDS up to the age of 75 years.8 Furthermore, several studies have suggested that allogeneic HCT is a viable option for patients aged 60 or older.9-13 Indeed, a remarkable change in patients' characteristics over time in our cohort was the progressive increase in older patients. In particular, the proportion of patients aged 60 years older drastically increased from 1.9% in 1997 to 51.0% in 2020, and the absolute annual numbers of HCT for age ≥70 years also increased over time, from 2 (1.6%) in 2000 to 34 (7.0%) in 2020 (Figure S3A), which could have contributed to no improvement of unadjusted OS and DFS over time. Although previous studies showed that chronological age did not significantly affect post-transplant survival of adult patients including those with MDS,9-11 our study demonstrated that older patients aged ≥60 years had significantly poorer transplant outcomes, whereas younger patients aged <40 years had significantly better transplant outcomes. Thus, chronological age remains an important prognostic factor for allogeneic HCT for adult patients with MDS. However, recent encouraging reports have demonstrated the efficacy and safety of allogeneic HCT in selected patients aged ≥70 years.13 Therefore, chronological age alone should not exclude an elderly patient as a candidate for allogeneic HCT, but optimal strategies for pretransplant risk assessment in elderly patients, such as geriatric assessment, will be needed for the better selection of older adults with MDS. The marked increase in the number of UBMT and UCBT in our cohort could have overcome the donor unavailability for MDS patients to facilitate allogeneic HCT, particularly in the latter two time periods. Although survival was significantly worse in patients who received UBMT and UCBT compared to those who received related BMT during the entire study period in the multivariate analysis, OS was improved in the latter two time periods (2001–2010, 2011–2020) than in the earlier period (1987–2000) among patients who received UBMT. Among patients who received UCBT, a drastic improvement of OS was observed in the most recent time period (2011–2020) compared to those in the middle time period (2001–2010), which is consistent with our previous reports of UCBT for adult AML and various hematological diseases.3, 4 Surely, matched sibling donor (MSD) has remained a frontline donor source followed by matched unrelated donor or unrelated cord blood for MDS, but a recent study by the European Group for Blood and Marrow Transplantation demonstrated higher DFS after allogeneic HCT for MDS using younger matched unrelated donors compared with older MSDs.14 Furthermore, rapid availability is the most important benefit for alternative HLA mismatched donors, such as unrelated cord blood or a haploidentical related donor,15, 16 for MDS patients, because some patients needed an urgent allogeneic HCT (Figure S3B). This ease of donor availability could extend access to allogeneic HCT for MDS because the identification of an MSD can be limited for patients with MDS, particularly in the elderly. In conclusion, our data could provide valuable evidence for allogeneic HCT for MDS patients in a real-world setting. Survival and NRM have improved after allogeneic HCT for adult patients with MDS, but a reduction of relapse rate will remain a challenge for further improvement. We thank all of the physicians and staff at the centers who provided the clinical data to the Transplant Registry Unified Management Program of the Japanese Data Center for Hematopoietic Cell Transplantation and the Japanese Society for Transplantation and Cellular Therapy (JSTCT). 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, under Grant Number 18ek0510023h0002. The authors declare no competing financial interests, except that MS received research support from Nippon Shinyaku and owns stock in Celaid Therapeutics. The data that support the findings of this study are available from the corresponding author upon reasonable request. Figure S1. Figure S2. Figure S3. Supplementary Table S1. Patient characteristics. Supplementary Table S2. Multivariate analysis of overall mortality, treatment failure, relapse, and non-relapse mortality. Supplementary Table S3. Cause of death according to time periods. Supplementary Table S4. Subgroup analysis of adjusted hazard ratio of overall mortality and treatment failure. Data S1. Supporting Information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.