To the Editor,
Myelodysplastic syndromes (MDS) are haematological neoplasms primarily affecting older adults, with increasing incidence in ageing societies such as Japan. 1 , 2 Recent genomic analyses have yielded key insights, including clonal haematopoiesis of indeterminate potential, specific genomic mutations and the prognosis. 3 Several therapeutic agents, such as lenalidomide, 4 darbepoetin arfa, 5 azacitidine (AZA), 6 luspatercept 7 and deferasirox, 8 have become available in Japan, along with improved antibiotics, antifungal agents and supportive care. 9 , 10 Improvements in allogeneic haematopoietic stem cell transplantation (allo‐HSCT) have also enhanced outcomes through expanded donor availability, conditioning regimens and improved prophylaxis and treatment of graft‐versus‐host disease. 11 AZA has been shown to prolong overall survival (OS) in higher risk MDS patients, whereas allo‐HSCT remains the only curative option. 6 , 11 Despite these improvements, real‐world evaluations of the temporal changes in treatment outcomes in this context, particularly those encompassing large cohorts, remain limited. We have conducted a retrospective analysis of MDS in Nagasaki Prefecture; however, temporal changes in survival have not yet been evaluated. 12 Therefore, we retrospectively analysed survivals in MDS patients across Nagasaki Prefecture, Japan, a region with a population of over 1.3 million, over the past 20 years.
This retrospective study included patients aged ≥16 years who were diagnosed with MDS between January 2000 and December 2022 at the 10 institutions in Nagasaki Prefecture, establishing a Nagasaki‐MDS database. This study aimed to investigate improvements in real‐world survival outcomes in MDS patients over an approximately two‐decade period. Patients with de novo MDS, therapy‐related MDS and secondary MDS after aplastic anaemia were included, whereas those with myelodysplastic/myeloproliferative neoplasms, including chronic myelomonocytic leukaemia, were excluded. According to the Revised International Prognostic Scoring System (IPSS‐R), patient scores with ≤3.5 and >3.5 at diagnosis were categorised as lower and higher risk MDS (LR‐ and HR‐MDS). 13 , 14 Patients were divided into two groups: Group 1 (2000–2010) and Group 2 (2011–2022), based on the approval of AZA for the treatment of MDS in Japan in 2011. The data were updated in January 2024. Probability of OS and cumulative incidence of acute myeloid leukaemia (AML) transformation were analysed. Univariate and multivariate analyses were performed for OS and AML transformation by using the clinical variables. The adjusted probabilities of OS and cumulative incidence of AML transformation were estimated, with consideration of other significant variables in the multivariate analysis.
A total of 1587 patients were initially registered. A small number (13 patients: 12 in Group 1 and 1 in Group 2)were unable to undergo cytogenetic analysis due to bone marrow fibrosis or other reasons and therefore excluded, leaving a final database of 1574 patients. The study cohort consisted of 573 patients in Group 1 and 1001 patients in Group 2 (Table 1). Group 2 patients were significantly older, with more patients aged ≥80 years (p < 0.001) and showed male predominance (p = 0.008), a higher proportion of World Health Organization subtypes with elevated blasts (p < 0.001), poorer cytogenetic risk (p = 0.009) and higher IPSS‐R risk (p < 0.001). The rate of allo‐HSCT was similar between the two groups (4.5% in Group 1 and 5.1% in Group 2; p = 0.710). Group 2 included more patients in the very poor cytogenetic risk subgroup by IPSS‐R and fewer with a normal karyotype (p = 0.009, Table S1). In Group 1, 33 patients (26 in LR‐MDS, 7 in HR‐MDS) received AZA after disease progression that occurred in or after 2011.
TABLE 1.
Comparison of patient characteristics between the two diagnosed era groups.
| Parameter | Group 1 (2000–2010) | Group 2 (2011–2022) | p‐Value |
|---|---|---|---|
| No. of patients | 573 | 1001 | |
| Age, median (range) | 73 (18–96) | 76 (16–96) | <0.001 a |
| Age group | |||
| <60 years, n (%) | 108 (18.9) | 88 (8.8) | <0.001 b |
| 60–69 years, n (%) | 113 (19.7) | 182 (18.2) | |
| 70–79 years, n (%) | 218 (38.0) | 363 (36.3) | |
| ≥80 years, n (%) | 134 (23.4) | 368 (36.8) | |
| Sex | |||
| Male, n (%) | 325 (56.7) | 637 (63.6) | 0.008 b |
| Female, n (%) | 248 (43.3) | 364 (36.4) | |
| Therapy‐related MDS | |||
| Yes, n (%) | 42 (7.3) | 102 (10.2) | 0.071 b |
| No, n (%) | 531 (92.7) | 899 (89.8) | |
| Progression from aplastic anaemia | |||
| Yes, n (%) | 10 (1.7) | 10 (1.0) | 0.299 b |
| No, n (%) | 563 (98.3) | 991 (99.0) | |
| MDS with paroxysmal nocturnal haemoglobinuria | |||
| Yes, n (%) | 3 (0.5) | 7 (0.7) | 0.926 b |
| No, n (%) | 570 (99.5) | 994 (99.3) | |
| WHO 2016 classification | |||
| MDS‐SLD, n (%) | 53 (9.2) | 107 (10.7) | <0.001 b |
| MDS‐MLD, n (%) | 348 (60.7) | 438 (43.8) | |
| MDS‐RS‐SLD, n (%) | 2 (0.3) | 13 (1.3) | |
| MDS‐RS‐MLD, n (%) | 7 (1.2) | 26 (2.6) | |
| MDS with isolated del(5q), n (%) | 7 (1.2) | 15 (1.5) | |
| MDS‐EB‐1, n (%) | 66 (11.5) | 169 (16.9) | |
| MDS‐EB‐2, n (%) | 53 (9.2) | 155 (15.5) | |
| MDS‐U, n (%) | 37 (6.5) | 78 (7.8) | |
| MDS‐U with 1% blood blasts, n (%) | 10 (1.7) | 23 (2.3) | |
| MDS‐U with SLD and pancytopenia, n (%) | 23 (4.0) | 33 (3.3) | |
| MDS‐U based on defining cytogenetic abnormality, n (%) | 4 (0.7) | 22 (2.2) | |
| IPSS‐R | |||
| Very low, n (%) | 57 (9.9) | 64 (6.4) | <0.001 b |
| Low, n (%) | 216 (37.7) | 306 (30.6) | |
| Intermediate, n (%) | 154 (26.9) | 270 (27.0) | |
| High, n (%) | 77 (13.4) | 164 (16.4) | |
| Very high, n (%) | 69 (12.0) | 197 (19.7) | |
| Cytogenetics risk by IPSS‐R | |||
| Very good, n (%) | 25 (4.4) | 39 (3.9) | 0.009 b yy |
| Good, n (%) | 347 (60.6) | 529 (52.8) | |
| Intermediate, n (%) | 94 (16.4) | 175 (17.5) | |
| Poor, n (%) | 38 (6.6) | 76 (7.6) | |
| Very poor, n (%) | 69 (12.0) | 182 (18.2) | |
| Received allo‐HSCT | |||
| Yes, n (%) | 26 (4.5) | 51 (5.1) | 0.710 b |
| No, n (%) | 547 (95.5) | 950 (94.9) | |
Abbreviations: allo‐HSCT, allogeneic haematopoietic stem cell transplantation; IPSS‐R, revised international prognostic scoring system; MDS, myelodysplastic syndromes; MDS‐EB‐1, myelodysplastic syndrome with excess blasts 1; MDS‐EB2, myelodysplastic syndrome with excess blasts 2; MDS‐MLD, myelodysplastic syndrome with multilineage dysplasia; MDS‐RS‐SLD, myelodysplastic syndrome with ring sideroblasts and single lineage dysplasia; MDS‐SLD, myelodysplastic syndrome with single lineage dysplasia; MDS‐U, myelodysplastic syndrome, unclassifiable; WHO, World Health Organization.
Calculated using Mann–Whitney U‐test.
Calculated using Pearson's chi‐squared test.
Table S2 presents the characteristics of LR‐MDS patients in Groups 1 and 2. Group 2 patients were significantly older, with more patients aged ≥80 years, and had a higher proportion of Bone Marrow (BM) blasts, while other variables showed no significant differences. The unadjusted 2‐year OS rates were higher in Group 1 than in Group 2: 81.7% (95% confidence interval [CI]: 76.9–85.6) versus 76.2% (95% CI: 71.7–80.1; p = 0.011) (Figure 1A). Univariate and multivariate analyses for OS are shown in Tables S3 and S4. In multivariate analysis, the following variables were associated with OS: age group, sex, absolute neutrophil count (ANC), Hb level, Platelet (Plt) count, percentage of BM blasts and karyotype risk. In addition, the era of diagnosis had no independent effect on OS (hazard ratio [HR]: 1.00 [95% CI: 0.80–1.26; p = 0.981]). The adjusted OS curves for the two groups largely overlapped, showing no meaningful separation between Groups 1 and 2 (Figure 1B), with the adjusted 2‐year OS rate of 81.6% (95% CI: 77.3–86.2) in Group 1 and 82.2% (95% CI: 78.6–86.0) in Group 2. The unadjusted cumulative incidence curves of AML transformation were almost overlapping, with 4.2% in Group 1 and 3.5% in Group 2 at 2 years (p = 0.938, Figure 1C). Univariate and multivariate analyses were performed (Tables S3 and S4). In the multivariate analysis, age group, percentage of BM blasts and karyotype risk were associated with AML transformation. The adjusted cumulative incidence curves were similar to the unadjusted curves, with 2.8% (95% CI: 1.5–5.3) in Group 1 and 2.2% (95% CI: 1.2–3.9) in Group 2 at 2 years (Figure 1D). No significant differences in the cause of death were observed between the groups (Figure S1).
FIGURE 1.
Probability of OS and cumulative incidence of AML transformation in patients with MDS. (A) Unadjusted and (B) adjusted OS curves in patients with LR‐MDS in Groups 1 and 2. (C) Unadjusted and (D) adjusted cumulative incidence curves of AML transformation in patients with LR‐MDS in Groups 1 and 2. (E) Unadjusted and (F) adjusted OS curves in patients with HR‐MDS in Groups 1 and 2. (G) Unadjusted and (H) adjusted cumulative incidence curves of AML transformation in patients with HR‐MDS in Groups 1 and 2. Adjusted survival curves reveal no significant difference in patients with LR‐MDS; however, a clear difference persists in patients with HR‐MDS. Regarding the cumulative incidence, significant differences were not observed in either unadjusted or adjusted analyses. allo‐HSCT, allogeneic haematopoietic stem cell transplantation; AML, acute myeloid leukaemia; AZA, azacitidine; HR, higher risk; LR, lower risk; MDS, myelodysplastic syndrome; OS, overall survival.
Table S5 presents the characteristics of HR‐MDS patients in Groups 1 and 2. Significant differences were observed in the median age and the distribution of age groups. Table S6 summarises the treatment strategies in the two groups. In Group 2, 42% (218 of 519 patients) received AZA, including some in clinical trials, with a median interval from diagnosis to AZA initiation of 33 days (range: 3–2669). Ten patients received other treatments before AZA (anabolic steroids, n = 3; darbepoetin alpha, n = 3; ciclosporin, n = 3; lenalidomide, n = 1). In contrast, approximately 90% of Group 1 patients did not receive chemotherapy. Among Group 2 patients, more than half of those aged <80 years received allo‐HSCT or chemotherapy (including AZA), compared with only approximately 20% of those aged ≥80 years (Figure S2). The unadjusted 2‐year OS rate was 37.4% (95% CI: 30.5–44.2) in Group 1 versus 41.6% (95% CI: 36.9–46.2) in Group 2 (p = 0.422, Figure 1E), showing no significant differences. Univariate and multivariate analyses for OS are shown in Tables S7 and S8. In the multivariate analysis, the following variables were significantly associated with OS: age group, sex, treatment‐related MDS, Hb level, Plt count, percentage of BM blasts and karyotype risk. While univariate analysis did not show a significant association, multivariate analysis revealed that the era of diagnosis was independently associated with OS (HR: 0.75 [95% CI: 0.62–0.91; p = 0.004, Table S8]). The adjusted OS curves are separated, with 29.9% (95% CI: 23.4–38.2) in Group 1 and 41.4% (95% CI: 36.3–46.7) in Group 2 at 2 years (Figure 1F). The unadjusted cumulative incidence curves of AML transformation were almost similar between two groups, with 2‐year cumulative incidence rates of 31.7% (95% CI: 25.3–38.3) in Group 1 and 31.6% (95% CI: 27.4–35.9) in Group 2 (p = 0.994, Figure 1G). Univariate and multivariate analyses were performed (Tables S7 and S8). In the multivariate analysis, the percentage of BM blasts was associated with AML transformation. The adjusted cumulative incidence curves were almost similar to the unadjusted curves, with 2‐year rates of 31.5% (95% CI: 25.4–38.7) in Group 1 and 28.6% (95% CI: 24.6–33.1) in Group 2 (Figure 1H). Censoring of allo‐HSCT did not substantially affect these results (data not shown).
In this study, we assessed era‐level changes in real‐world outcomes of 1574 MDS patients in Nagasaki Prefecture over two decades, without evaluating the comparative effectiveness of specific treatments. Although this database includes 127 atomic bomb survivors, survivor status did not affect the outcomes in the multivariable analysis of entire cohort (data not shown). Treatment allocation was non‐random and influenced by clinical judgement and patient factors. Survival significantly improved in HR‐MDS patients, while LR‐MDS outcomes remained unchanged, likely reflecting multiple factors such as advances in supportive care (such as anti‐infective therapies), introduction of novel agents (such as darbepoetin alpha and AZA) and refinements in allo‐HSCT. As this is a retrospective study including patients who were not eligible for clinical trials, our analysis might better capture the reality of clinical practice. Our analysis trends towards older age at diagnosis and increased MDS incidence, consistent with the predominance of MDS in the elderly, as previously reported. 2 LR‐MDS patients showed no improvement in OS or AML transformation, and the causes of death remained unchanged. Conversely, adjusted OS improved, though AML transformation rates did not in HR‐MDS patients, possibly due to limited AZA use in this older cohort. These findings differ somewhat from prior large‐scale analyses, potentially reflecting regional, temporal and demographic variations. 15 Limitations include incomplete clinical data (e.g. Eastern Cooperative Oncology Group [ECOG] performance status, lactate dehydrogenase [LDH]) and small subgroup sizes, limiting robust statistical adjustment (e.g. landmark analysis, propensity score methods). In addition, 33 patients (26 in LR‐MDS and 7 in HR‐MDS) received AZA after its approval, and a potential impact on survival could not be excluded. Thus, survival differences, particularly regarding AZA, may reflect confounding by indication rather than causal effects. In conclusion, while survival has improved in HR‐MDS, LR‐MDS remains a significant challenge. Progress in supportive care and specific treatments likely contributed to HR‐MDS outcomes, highlighting the need for expanded therapeutic options in the next decade.
AUTHOR CONTRIBUTIONS
TF, KA and Y Miyazaki were responsible for study conception, data analysis and manuscript preparation. ET, HT, HS, MC, HK, JN, TK, MT, MH, HI, SS, YS, DI, YI, TJ, SY and Y Moriuchi collected the clinical data. All the authors critically revised the manuscript and approved its final submission.
FUNDING INFORMATION
This study was supported in part by the Japan Society for the Promotion of Science KAKENHI under grants JP22K08481 (K. Ando) and 25K11628 (Y. Miyazaki) and by the Program of the Network‐Type Joint Usage/Research Center for Radiation Disaster Medical Science.
CONFLICT OF INTEREST STATEMENT
Y. Miyazaki received honoraria from Abbvie Co. Ltd., Astellas Co. Ltd., Chugai Co. Ltd., Novartis Co. Ltd., Bristol Myers Squib Co. Ltd. and Daiichi‐Sankyo Co. Ltd., and is an Associate Editor of the International Journal of Hematology. The other authors declare no conflicts of interest.
ETHICS STATEMENT
This study was conducted in accordance with the ethical standards of the World Medical Association's Declaration of Helsinki.
PATIENT CONSENT STATEMENT
This retrospective study used an opt‐out approach. In accordance with the institutional policy, informed consent was waived,and patients were given the opportunity to decline participation by accessing publicly available information.
CLINICAL TRIAL REGISTRATION
This study was approved by the Institutional Review Board of Nagasaki University Hospital (approval no. 19090906‐5) and the corresponding committee of each participating institute.
Supporting information
Figure S1.
Figure S2.
Tables S1–S8.
ACKNOWLEDGEMENTS
Data for this study were obtained from 10 institutions in Nagasaki, Japan: the NHO Nagasaki Medical Center (Drs. Nishimura, Sakaki, Kobayashi, Imaizumi and Yoshida), Sasebo City General Hospital (Drs. Ichinose, Fujioka, Kamijo, Makiyama and Moriuchi), Nagasaki Harbor Medical Center (Drs. Hata and Tsushima), Japanese Red Cross Nagasaki Genbaku Hospital (Drs. Matsuo, Sawayama, Taguchi and Jo), Japan Community Health Care Organization Isahaya General Hospital (Drs. Chiwata and Nakashima), Saint Francis Hospital (Dr. Takasaki), Nagasaki Prefecture Shimabara Hospital (Dr. Kitanosono), Nagasaki Prefecture Gotochuoh Hospital (Drs. Matsumoto and Imanishi), NHO Nagasaki Hospital (Dr. Kawaguchi) and Nagasaki University Hospital (Drs. Katsuoka, Hashimoto, Toriyama, Sakamoto, Kato, Taguchi, Horai, Itonaga and Sato).
DATA AVAILABILITY STATEMENT
Data supporting the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Garcia‐Manero G. Myelodysplastic syndromes: 2023 update on diagnosis, risk‐stratification, and management. Am J Hematol. 2023;98(8):1307–1325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Chihara D, Ito H, Katanoda K, Shibata A, Matsuda T, Sobue T, et al. Incidence of myelodysplastic syndrome in Japan. J Epidemiol. 2014;24(6):469–473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Cook MR, Karp JE, Lai C. The spectrum of genetic mutations in myelodysplastic syndrome: should we update prognostication? EJHaem. 2022;3(1):301–313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Fenaux P, Giagounidis A, Selleslag D, Beyne‐Rauzy O, Mufti G, Mittelman M, et al. A randomized phase 3 study of lenalidomide versus placebo in RBC transfusion‐dependent patients with low‐/Intermediate‐1‐risk myelodysplastic syndromes with del5q. Blood. 2011;118(14):3765–3776. [DOI] [PubMed] [Google Scholar]
- 5. Fenaux P, Santini V, Spiriti MAA, Giagounidis A, Schlag R, Radinoff A, et al. A phase 3 randomized, placebo‐controlled study assessing the efficacy and safety of epoetin‐alpha in anemic patients with low‐risk MDS. Leukemia. 2018;32(12):2648–2658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Fenaux P, Mufti GJ, Hellstrom‐Lindberg E, Santini V, Finelli C, Giagounidis A, et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher‐risk myelodysplastic syndromes: a randomised, open‐label, phase III study. Lancet Oncol. 2009;10(3):223–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Fenaux P, Platzbecker U, Mufti GJ, Garcia‐Manero G, Buckstein R, Santini V, et al. Luspatercept in patients with lower‐risk myelodysplastic syndromes. N Engl J Med. 2020;382(2):140–151. [DOI] [PubMed] [Google Scholar]
- 8. Yang S, Zhang MC, Leong R, Mbuagbaw L, Crowther M, Li A. Iron chelation therapy in patients with low‐ to intermediate‐risk myelodysplastic syndrome: a systematic review and meta‐analysis. Br J Haematol. 2022;197(1):e9–e11. [DOI] [PubMed] [Google Scholar]
- 9. Contejean A, Maillard A, Canouï E, Kernéis S, Fantin B, Bouscary D, et al. Advances in antibacterial treatment of adults with high‐risk febrile neutropenia. J Antimicrob Chemother. 2023;78(9):2109–2120. [DOI] [PubMed] [Google Scholar]
- 10. Wang J, Zhou M, Xu JY, Zhou RF, Chen B, Wan Y. Comparison of antifungal prophylaxis drugs in patients with hematological disease or undergoing hematopoietic stem cell transplantation: a systematic review and network meta‐analysis. JAMA Netw Open. 2020;3(10):e2017652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. DeFilipp Z, Ciurea SO, Cutler C, Robin M, Warlick ED, Nakamura R, et al. Hematopoietic cell transplantation in the management of myelodysplastic syndrome: an evidence‐based review from the American society for transplantation and cellular therapy committee on practice guidelines. Transplant Cell Ther. 2023;29(2):71–81. [DOI] [PubMed] [Google Scholar]
- 12. Toriyama E, Hata T, Yokota KI, Chiwata M, Kamijo R, Hashimoto M, et al. No clear survival benefit of azacitidine for lower‐risk myelodysplastic syndromes: a retrospective study of Nagasaki. Cancer Sci. 2020;111(12):4490–4499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Pfeilstöcker M, Tuechler H, Sanz G, Schanz J, Garcia‐Manero G, Solé F, et al. Time‐dependent changes in mortality and transformation risk in MDS. Blood. 2016;128(7):902–910. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. NCCN guidelines® insights: myelodysplastic syndromes, version 2.2025. J Natl Compr Cancer Netw. 2025;23(3):66–75. [DOI] [PubMed] [Google Scholar]
- 15. Neukirchen J, Nachtkamp K, Schemenau J, Aul C, Giagounidis A, Strupp C, et al. Change of prognosis of patients with myelodysplastic syndromes during the last 30 years. Leuk Res. 2015;39(7):679–683. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1.
Figure S2.
Tables S1–S8.
Data Availability Statement
Data supporting the findings of this study are available from the corresponding author upon reasonable request.