Myelodysplastic Syndrome Treatment Paradigm Comparisons: A Study of Japanese Real-World Health Claims Data

Abstract

Background: Complex pathophysiology and high incidence of myelodysplastic syndromes (MDS) limits treatment options. Azacitidine (AZA) is the standard first-line (1L) treatment for high-risk MDS in Japan. This study evaluated treatment patterns and overall survival (OS) of patients with MDS in a real-world setting in Japan.

Materials and methods: This non-interventional, retrospective study examined patients with MDS aged ≥ 18 years between April 2008 and September 2018. The primary outcome was the proportion of patients who received MDS-related treatment regimens as 1L in main cohort. Secondary outcomes included treatment patterns from 1L to third-line (3L) in main cohort and OS in 1L-AZA cohort (patients from main cohort who had AZA as 1L receiving 7-day or 5-day AZA dosing schedule).

Results: Of 6,204 eligible patients, 1,068 comprised 1L-AZA cohort. Most patients received supportive therapies (53.2%) as 1L treatment, followed by antineoplastic agents (39.5%), with 31% receiving AZA. Approximately, 47% of patients switched to second-line (2L) treatment, while 25.3% advanced to 3L treatment. Patients in 1L-AZA cohort receiving 5-day AZA was associated with longer OS compared to 7-day AZA (median OS: 38.4 vs. 20.4 months; p < 0.001).

Conclusion: The majority of Japanese patients with MDS received supportive therapies as 1L followed by AZA. Five-day AZA treatment significantly improved OS compared with 7-day AZA.

Graphical abstract

Keywords

myelodysplastic syndrome / azacitidine / supportive therapy / Japan / overall survival

Introduction

Myelodysplastic syndromes (MDS) are a group of heterogeneous disorders originating from hematopoietic stem cells characterized by ineffective hematopoiesis, bone marrow dysplasia, peripheral blood cytopenias and are associated with an increased risk of progression to acute myeloid leukemia (AML)[1,2]. The risk of MDS increases with age, and the median age of patients at the time of diagnosis is approximately 70 years[3]. The overall prognosis of MDS is poor, and the only potential curative treatment is limited to allogeneic stem cell transplantation (alloSCT)[4]. Inherent complex pathophysiology of MDS and increased incidence in older population often limits available treatment options[5].

In Japan, azacitidine (AZA) is the only first line (1L) of treatment available in elderly patients with high-risk MDS (HR-MDS) and who are ineligible for alloSCT. The Japanese Society of Hematology guidelines also recommend lenalidomide in MDS patients with deletion of the long arm of chromosome 5 (del^[5q]) and chemotherapy in patients with MDS who are intolerant to AZA[6]. In a pivotal AZA study (AZA-001 trial), treatment with AZA in a 28-day cycle at 75 mg/m² per day for seven consecutive days improved overall survival compared with conventional care regimens in patients with HR-MDS[7]. Due to standard clinical practice, most MDS patients in Japan are expected to be HR-MDS, although some may include low-risk MDS (LR-MDS) as AZA is approved for all-type MDS in Japan.

The preliminary feasibility assessment showed that around 41% of HR-MDS patients on AZA monotherapy terminated their treatment within 4 cycles[8], which is substantially lower compared with a median of 9 cycles in AZA-001 study[7]. Given the contrasting outcomes observed in results from clinical studies and meta-analyses[9,10], it is difficult to benchmark the patient responses and outcomes in patients treated with AZA. Therefore, a detailed understanding of the current real-world treatment patterns in patients with MDS is important to identify unmet needs and help in addressing them for newer treatment options. The overall objective of this study was to describe treatment patterns and overall survival (OS), in a real-world cohort of patients with MDS in Japan.

Materials and methods

Study design

This was a non-interventional, retrospective cohort study of patients with MDS using data from the Medical Data Vision Co., Ltd (MDV; Tokyo, Japan) database in Japan. The overall study period (1 April 2008 to 30 September 2018) included pre- and post-index periods, where the index date was defined as the date of first confirmed diagnosis of MDS, as defined by the International Classification of Diseases-Clinical Modification (ICD-10-CM) code: D46, during the identification period (1 October 2008 to 30 September 2017). The baseline period was 6 months preceding the index date and the follow-up period spanned from the index date to end of data availability or patients’ death (supplementary Figure 1).

Data source and ethics

The study used MDV database which is a health claims database in Japan that accumulates data from over 480 hospitals and all insurance types and includes anonymized inpatient and outpatient data and laboratory test results for over 44 million patients[11]. The primary strengths of the MDV database lie in its extensive sample size, and the inclusion of laboratory test results. The database is accessible to private companies and do not require any Institutional Review Board submission for data access[12].

Study population

Patients included in the main cohort were aged 18 years or older with an inpatient diagnosis of MDS or 2 outpatient confirmed diagnoses of MDS during the identification period, had bone marrow biopsy/aspiration procedures in ± 3 months from index date, and had at least 1 MDS-related treatment after index date (Figure 1). MDS-related treatments administered as 1L are presented in Figure 2. Further, patients from the main cohort who have received AZA as 1L treatment with a 5-day or 7-day AZA dosing schedule only were referred to as 1L-AZA cohort. Patients who received other or mixed AZA dosing schedules were excluded from this cohort. Patients were excluded from the study if (i) they had any history of SCT (Remuneration code: K922), pre-SCT conditioning (ATC: L01A0), graft versus host disease (ICD-10-CM code: D89.81), post-transplant lymphoproliferative disorder (ICD-10-CM code: D47.Z1) or complications of transplanted organ (ICD-10-CM code: T86) within 12 months prior to the index date; (ii) they had confirmed AML (ICD-10-CM codes: C920, C927, C929) diagnosis on index date or at any time point prior to the index date.

Treatment schedule for azacitidine in 1L-AZA cohort

Azacitidine was administered as per the following schedule in a 28-day cycle: (i) 7-0-0 (for seven consecutive days); (ii) 6-1-1 (for six consecutive days, followed by a 1-day off, and then continued for 1 day); (iii) 5-2-2 (for five consecutive days, followed by a 2-day off, and then continued for 2 days); (iv) 5-0-0 (for five consecutive days). The 7-day AZA dosing schedule included the 7-0-0, 6-1-1 and 5-2-2 administration schedules; 5-day AZA dosing schedule included the 5-0-0 administration schedule.

Outcomes

The primary outcome was the proportion of patients who received MDS-related treatment regimens as 1L of treatment in the main cohort. Secondary outcomes included, baseline demographic and clinical characteristics; treatment patterns from 1L to third line (3L) of treatment in the main cohort; overall survival (OS) among MDS patients in 1L-AZA cohort with different AZA treatment administration schedules (7-day AZA dosing schedule or 5-day AZA dosing schedule) and by different age groups (< 70 and ≥ 70 years); association between potential predictive factors and OS of patients with MDS in 1L-AZA cohort; proportion of patients receiving different AZA treatment administration schedules categorized by number of treatment cycles (1–3, 4–6, 7–9, and > 9 cycles) in 1L-AZA cohort.

Statistical analyses

For all descriptive statistics analyses, counts and percentages were presented for categorical variables; measures of central tendency (mean and median) together with standard deviations were used to summarize the continuous variables. A Sankey diagram was created to visualize treatment-switching patterns over the first three lines of treatment. The Sankey diagram followed the priority rule to categorize prescribed treatments appropriately. The 1L of treatment was considered as primary and if two treatments were prescribed on the very first record, the priority was antineoplastic agents > all other treatments > supportive treatments, resulting in only one treatment per patient for each line represented in the diagram.

Overall survival in months was estimated using Kaplan–Meier method with different AZA administration schedules (7-day AZA dosing schedule or 5-day AZA dosing schedule) and the OS estimates were stratified by age. Univariate and multivariate Cox proportional hazards models were used to examine the association between potential predictive factors and OS among patients with MDS in 1L-AZA cohort. In the univariate analysis, hazard ratio (HR) and 95% confidence interval (CI) were estimated for each predictor variable. Variables that achieve statistical significance (p ≤ 0.05) in the univariate analysis were included in the multivariate model. The proportional hazards assumption was evaluated using Schoenfeld residuals and log-minus-log plots. Variables violating the proportional hazards assumption were incorporated as stratification factors in a stratified Cox proportional hazards model. The final multivariable model included AZA dosing schedule and age group as covariates, while the following variables were included as stratification factors due to violation of the proportional hazards assumption: time to AZA treatment initiation (within 1 month/beyond 1 month), number of AZA treatment cycle, infection related drug use < 3 months of AZA initiation, bleeding events or occurrence of platelet transfusion < 3 months of AZA initiation. Adjusted HR (aHR) with corresponding 95% CI and p-value are reported. Chi-square test was used to compare the proportion of patients in AZA treatment administration schedules (7-day AZA dosing schedule or 5-day AZA dosing schedule). All statistical analyses were performed using R version 4.0.2.

Results

Overall, 30,808 patients with MDS during the identification period were extracted from the MDV database. Of these, 6,204 (20.1%) patients met all the inclusion and exclusion criteria and were included in this study as the main cohort (Figure 1). Of the patients in main cohort, 1,068 (17.2%) patients were included in 1L-AZA cohort.

Patient characteristics

The baseline demographic and clinical characteristics of patients with MDS at index date in main cohort and 1L-AZA cohort are presented in Table 1. In the main cohort, the mean age of patients was 72.4 years (median: 74 years) with majority of the patients aged ≥ 60 years (87.8%) and mean baseline Charlson Comorbidity Index score of 1.7 (median: 1.0). The proportion of transfusion-dependent patients in the main cohort was 9.4% and 0.6% of patients were having the iron overload.

MDS-related treatment regimens as first line of treatment

The proportion of patients with MDS in main cohort receiving different MDS-related treatment regimens as 1L of treatment after index date is presented in Figure 2. Of 6,204 patients in the main cohort, the most common treatment regimen received as 1L of treatment was supportive therapies (53.2%), followed by antineoplastic agents (39.5%) and immunosuppressive agents (26.3%). Among the antineoplastic agents, AZA was the most common treatment. In patients receiving 7-day AZA the most preferred schedule was 7-0-0, being used in 11.1% of patients, followed by 5-2-2 schedule in 4.0% of patients and 0.4% of patients received the 6-1-1 schedule. The 5-day AZA dosing schedule (5-0-0) was used in 11.1% of patients.

Treatment patterns

Changes in treatment regimens across 1L, 2L, and 3L of MDS-related treatments in main cohort are presented in online supplementary Figure 2. Approximately 40% of patients were treated with antineoplastic agents as their 1L of treatment, with AZA being the most common 1L antineoplastic agent (31.1%). One-third (33.3%) of the patients were treated with 1L supportive therapy. It was observed that around 47% of patients switched to 2L of treatment, while a quarter (25.3%) of them advanced to 3L of treatment. Of the patients who switched from 1L to 2L of treatment, 12.7% of patients received supportive therapy, 20% of patients received antineoplastic agents, with AZA being the most common (10.2%) followed by chemotherapy (7.5%), and more than half (53%) of the patients had no mention of predefined drug. Of the patients who switched from 2L to 3L of treatment, a three-fourth (75%) of the patients had no mention of predefined drug (online supplementary Figure 2).

Overall survival

Kaplan–Meier curve for OS in patients in 1L-AZA cohort managed with 5-day AZA dosing schedule was associated with a longer OS compared to patients receiving 7-day AZA dosing schedule (median OS: 38.4 vs. 20.4 months; p < 0.001; Figure 3). The Kaplan–Meier curves for 1L-AZA cohort patients stratified by different age groups (< 70 and ≥ 70 years) are presented in online supplementary Figure 3. Among patients aged < 70 years, significant differences were observed in OS for different AZA treatment cycles (p < 0.001 by log-rank test). The greater number of AZA treatment cycles was associated with improved OS. Similar findings were also observed among patients aged ≥ 70 years (online supplementary Figure 3). In the multivariable stratified Cox proportional hazards model among MDS patients in the 1L-AZA cohort, older age (≥ 75 years; aHR: 1.70; 95% CI: 1.41, 2.05; p < 0.001) was independently associated with poorer survival, whereas a 5-day AZA dosing schedule (5-0-0) was associated with improved survival compared with the 7-day schedule (7-0-0) (aHR: 0.66; 95% CI: 0.55, 0.80; p < 0.001). These potential predictors of OS are presented in Table 2.

The proportions of patients with different AZA treatment administration schedules (7-day AZA dosing schedule vs. 5-day AZA dosing schedule) categorized by the number of treatment cycles (1–3, 4–6, 7–9, > 9 cycles) in 1L-AZA cohort are presented in online supplementary Table 1.

Discussion

This non-interventional, retrospective, real world evidence study assessed the outcomes of 6,204 patients with MDS in Japan. The study provides a current snapshot of real-world treatment patterns and OS in patients with MDS. Most patients received supportive therapies (53.2%) as 1L of treatment followed by antineoplastic agents (39.5%) with AZA being the most common treatment received in 31% of patients in the main cohort. This is consistent with the current clinical practice where AZA is the standard of care therapy for patients with MDS who are ineligible for alloSCT and intensive chemotherapies[6]. MDS is characterized as disease of elderly population, and this is also reflected in our study where the majority of patients (65.1%) were aged > 70 years. Given that the majority of patients are likely ineligible for alloSCT due to advanced age and the relatively limited treatment options approved in Japan[6], as compared with United States[13] and Europe[14], there is an urgent need for introducing new and advanced treatments. This would improve treatment strategies for MDS and facilitate the evaluation of potential benefits offered by new treatment options.

Interestingly, this study revealed that MDS patients who received 5-day AZA dosing schedule were associated with longer OS compared to those on a 7-day AZA dosing schedule (38.4 vs. 20.4 months) regardless of age. In Japan’s real-world clinical practice, among patients who underwent ≥ 4 cycles of AZA (4–6, 7–9, and > 9 cycles), a notably large proportion received the 5-day dosing schedule as opposed to the officially approved 7-day regimen (online supplementary Table 1). This trend could be attributed to the improved survival outcomes observed with the 5-day AZA dosing schedule compared to the 7-day AZA dosing schedule in patients with MDS, or might be due to a higher incidence of adverse events or greater risk among patients in the 7-day AZA dosing schedule group relative to those in the 5-day AZA dosing schedule group. This is the first study to suggest that a certain proportion of patients may benefit from a shorter, specifically a 5-day AZA dosing schedule. Miyazaki et al. attempted to answer this question with an RCT, however, their study lacked the statistical power necessary to conclusively establish the superiority of 7-day AZA dosing schedule over 5-day AZA dosing schedule with respect to OS[15]. Therefore, a clinical study comparing 5-day and 7-day AZA dosing schedule evaluating OS may be helpful in confirming the effective AZA dosing schedule. Unknown ethno-physiological factors could be at play and future pivotal trials for MDS should consider the 5-day AZA dosing schedule if this could be important in determining the treatment benefit of new investigational drugs among East Asian patients.

Survival outcomes improved with a greater number of AZA treatment cycles, with patients receiving > 9 cycles exhibiting the longest survival of 40.8 months[8]. This finding is in line with a retrospective study conducted using the Surveillance, Epidemiology and End Results (SEER)-Medicare database, where patients receiving ≥ 4 cycles of hypomethylating agents had a statistically significantly better median OS compared to patients receiving < 4 cycles (median OS: 16 months vs 4 months[16]. Several population-based studies from United States and Europe have showed similar results reporting median OS ranging from 11 to 17 months[17-19]. However, for the first time an improved OS with a greater number of cycles has been demonstrated in Japanese population in the current study. Our study also suggested that > 9 cycles of AZA may improve leukemia-free survival in MDS patients more than 70 years of age (online supplementary Figure 4). Further corroborating the benefit of more AZA cycles, about 27% patients with HR-MDS received > 9 cycles of AZA in this study[8], and a positive correlation was found between AZA treatment duration and patient survival.

Evidence from the existing literature suggests that there is a substantial economic burden, especially in patients who showed suboptimal response to AZA and require more supportive therapies[20-22]. The current study has also shown higher costs associated with AZA treatment in Japan. It is worth mentioning that patients who received < 4 cycles of treatment incurred higher total treatment costs compared to patients who received ≥ 4 cycles (4–6, 7–9, and > 9 cycles). The greater number of cycles did not translate into increased healthcare costs with lowest total treatment cost per patient per month in patients who received > 9 cycles[8]. These findings are similar to the previous studies that reported lower treatment cost per patient per month in patients who received more cycles of AZA treatment[21,22].

The generalizability of the results is shown by the better OS in patients receiving > 9 cycles of AZA, in alignment with the AZA-001 pivotal study. Future research should focus on developing new and advanced treatments, either alone or in combination with AZA, especially in the Japanese population that would reduce the rate of treatment failure and increase the overall response.

In real-world setting involving Japanese patients, higher proportion of elderly patients may be treated using AZA compared with pivotal clinical trials. The lower GI and hematopoietic side-effects of 5-day AZA may increase the long-term tolerability of treatment[6,23,24] in such vulnerable populations, despite the comparable pharmacokinetics/pharmacodynamics between Japanese and Caucasian patients[25]. Altogether, this may have contributed to the better OS amongst Japanese higher-risk MDS patients treated using 5-day AZA according to real-world data. Notably, this claims database analysis is severely limited by the lack of dose administration record, therefore further research is needed to confirm the association of survival benefit with relative dose intensity of AZA.

The study has several limitations. First, the MDS patients in the MDV database may not be representative of the general MDS population in Japan as MDV database only covers around 23% of acute hospitals in Japan. Second, while claims databases capture diagnostic and procedure codes, they lack information on disease severity necessary for patient classification based on specific scoring systems. A third limitation involves the common limitations inherent to claims databases such as potential coding errors or data omissions. Finally, only mortality within the inpatient setting could be reliably accounted.

Conclusion

This real-world observational study, utilizing claims data from Japanese patients with MDS, found that the majority of patients were treated with supportive therapies as first line of treatment followed by AZA. Additionally, this study revealed that MDS patients who received 5-day AZA dosing schedule had significantly improved OS compared to those on a 7-day AZA dosing schedule. However, due to the retrospective nature of the study, definite conclusions are limited. These findings highlight the need for innovative research designs to address the considerable unmet medical needs in patients with MDS.

References

[1] Arber DA, Orazi A, Hasserjian R, et al. The 2016 revision to the world health organization classification of myeloid neoplasms and acute leukemia. Blood. 2016;127(20):2391-2405.

[2] Cazzola M. Myelodysplastic syndromes. N Engl J Med. 2020;383(14):1358-1374.

[3] Surveillance Research Program, National Cancer Institute. All cancer sites combined: recent trends in SEER age-adjusted incidence rates, 2000-2019 by sex, delay-adjusted SEER incidence rate, all races, all ages. National Cancer Institute. Accessed October 26, 2024.

[4] Sekeres MA, Taylor J. Diagnosis and treatment of myelodysplastic syndromes: a review. JAMA. 2022;328(9):872-880.

[5] Aleshin A, Greenberg PL. Molecular pathophysiology of the myelodysplastic syndromes: insights for targeted therapy. Blood Adv. 2018;2(20):2787-2797.

[6] Miyazaki Y. JSH practical guidelines for hematological malignancies, 2018: i. leukemia-6 myelodysplastic syndromes (mds). Int J Hematol. 2020;111(4):481-493.

[7] Fenaux P, Mufti GJ, Hellstrom-Lindberg E, 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.

[8] Obara N, Shioi H, Ooi SH, et al. Real-world treatment patterns, outcomes and disease burden of japanese patients with high-risk myelodysplastic syndrome: a claims database study. Blood. 2022;140(Suppl 1):12327-12328.

[9] Hasegawa K, Wei AH, Garcia-Manero G, et al. Azacitidine monotherapy in patients with treatment-naïve higher-risk myelodysplastic syndrome: a systematic literature review and meta-analysis. Clin Lymphoma Myeloma Leuk. 2023;23(2):127-137.

[10] Garcia JS, Swords RT, Roboz GJ, et al. A systematic review of higher-risk myelodysplastic syndromes clinical trials to determine the benchmark of azacitidine and explore alternative endpoints for overall survival. Leuk Res. 2021;104:106555.

[11] MDV Database. Medical Data Vision Co., Ltd. Accessed November 29, 2024.

[12] Laurent T, Simeone J, Kuwatsuru R, et al. Context and considerations for use of two japanese real-world databases in japan: medical data vision and japanese medical data center. Drugs Real World Outcomes. 2022;9(2):175-187.

[13] Garcia-Manero G. Myelodysplastic syndromes: 2023 update on diagnosis, risk-stratification, and management. Am J Hematol. 2023;98(8):1307-1325.

[14] Fenaux P, Haase D, Santini V, et al. Myelodysplastic syndromes: esmo clinical practice guidelines for diagnosis, treatment and follow-up^†☆. Ann Oncol. 2021;32(2):142-156.

[15] Miyazaki Y, Kiguchi T, Sato S, et al. Prospective comparison of 5- and 7-day administration of azacitidine for myelodysplastic syndromes: a jalsg mds212 trial. Int J Hematol. 2022;116(2):228-238.

[16] Zeidan AM, Hu X, Zhu W, et al. Association of provider experience and clinical outcomes in patients with myelodysplastic syndromes receiving hypomethylating agents. Leuk Lymphoma. 2020;61(2):397-408.

[17] Zeidan AM, Stahl M, Sekeres MA, Steensma DP, Komrokji RS, Gore SD. A call for action: increasing enrollment of untreated patients with higher-risk myelodysplastic syndromes in first-line clinical trials. Cancer. 2017;123(19):3662-3672.

[18] Bernal T, Martínez-Camblor P, Sánchez-García J, et al. Effectiveness of azacitidine in unselected high-risk myelodysplastic syndromes: results from the spanish registry. Leukemia. 2015;29(9):1875-1881.

[19] Zeidan AM, Sekeres MA, Garcia-Manero G, et al. Comparison of risk stratification tools in predicting outcomes of patients with higher-risk myelodysplastic syndromes treated with azanucleosides. Leukemia. 2016;30(3):649-657.

[20] Zeidan AM, Wang R, Davidoff AJ, et al. Disease-related costs of care and survival among medicare-enrolled patients with myelodysplastic syndromes. Cancer. 2016;122(10):1598-607.

[21] Stein EM, Bonifacio G, Latrémouille-Viau D, et al. Healthcare resource utilization and costs in patients with myelodysplastic syndromes treated with hypomethylating agents: a seer-medicare analysis. J Med Econ. 2021;24(1):234-243.

[22] Joshi N, Kale H, Corman S, Wert T, Hill K, Zeidan AM. Direct medical costs associated with treatment nonpersistence in patients with higher-risk myelodysplastic syndromes receiving hypomethylating agents: a large retrospective cohort analysis. Clin Lymphoma Myeloma Leuk. 2021;21(3):e248-e254.

[23] Park S, Park SY, Lee JH, et al. Five-day versus 7-day treatment regimen with azacitidine in lower risk myelodysplastic syndrome: a phase 2, multicenter, randomized trial. Cancer. 2022;128(23):4095-4108.

[24] Sadashiv SK, Hilton C, Khan C, et al. Efficacy and tolerability of treatment with azacitidine for 5 days in elderly patients with acute myeloid leukemia. Cancer Med. 2014;3(6):1570-1578.

[25] Uchida T, Ogawa Y, Kobayashi Y, et al. Phase I and II study of azacitidine in japanese patients with myelodysplastic syndromes. Cancer Sci. 2011;102(9):1680-1686.

RIGHTS & PERMISSIONS

The Author(s) 2026. This article is published by Higher Education Press at journal.hep.com.cn.