Decitabine/Cedazuridine in the Management of Myelodysplastic Syndrome and Chronic Myelomonocytic Leukemia in Canada

The management of myelodysplastic syndrome (MDS) and chronic myelomonocytic leukemia (CMML) is limited and remains an unmet need. Decitabine/cedazuridine (DEC-C, ASTX727) is Canada’s first and only approved oral hypomethylating agent for MDS and CMML. We characterized the real-world use of DEC-C through a Canadian compassionate use program. Demographic and clinical data from 769 patients enrolled in Taiho Pharma Canada’s Patient Support Program were collected and analyzed. These patients represent a collection period from 10 November 2020 to 31 August 2022 with a median age of 76 years. Among 651 patients who started DEC-C, the median treatment duration was 4.2 cycles. The median overall and progression-free survival were 21.6 and 10.7 months, respectively. Among 427 patients who discontinued treatment, the majority (69.5%) stopped due to death (n = 164) or disease progression (n = 133). Multivariable cox regression showed that age, province of residence, blast counts, antibiotic prophylaxis, and number of dose reductions and delays were not significantly associated with overall and progression-free survival. DEC-C is a promising alternative to parenteral hypomethylating agent therapy, and it likely addresses an important unmet need for effective and convenient therapies in this setting.


Introduction
Myelodysplastic syndrome (MDS) encompasses a heterogeneous group of hematopoietic stem cell disorders characterized by ineffective hematopoiesis and dysplastic changes in myeloid, erythroid, and megakaryocytic progenitors [1]. The resulting pancytopenia and increased risk of progression to acute myeloid leukemia (AML) cause significant morbidity and mortality in patients with MDS [2]. Chronic myelomonocytic leukemia (CMML) is an overlapping myelodysplastic/myeloproliferative malignancy [3]. In Canada, the estimated age-standardized incidence of MDS is 3.69 per 100,000 [4]. CMML is less common, with an average incidence rate of 2.45 cases per million individuals annually [5]. While long-term epidemiological studies in Canada remain limited, yearly incidence rates have increased in several countries including Germany, the UK, and the United States [6]. In combination with increasing physician awareness and improved diagnostic classifications, it is likely that MDS in Canada's aging population will demonstrate similar trends in years to come [4].
With a median age of diagnosis between 70 and 76 years [2,4,7], most MDS patients are elderly with a varied prognosis depending on disease-related factors [8]. Patients with MDS may be classified according to the International Prognostic Scoring System (IPSS), by

Statistical Analysis
Descriptive statistics were used to summarize baseline patient and treatment characteristics. Continuous variables were presented as medians with interquartile ranges and means with standard deviations while categorical variables were presented as frequencies and percentages. Subgroup analysis was performed to analyze potential relationships between clinical parameters. Patients were grouped as LR-MDS (IPSS status intermediate- 1) or HR-MDS (IPSS status intermediate-2 and high risk) as per IPSS [9]. Further analysis explored parameters based on treatment duration. Subgroups were divided between patients receiving greater than, or equal to, 4 cycles of DEC-C versus less than 4 cycles. Each cycle is defined in 28-day periods as per the product monograph [22]. The 6th cycle re-enrollment data were also utilized to compare the transfusion status against initial enrollment forms to note any change in the last 6 cycles.
Progression-free survival (PFS) and overall survival (OS) were estimated using the Kaplan-Meier method and utilizing time-to-event data involving date of enrollment to date of event (i.e., treatment discontinuation due to death or disease progression).
Wilcoxon rank-sum tests were used for continuous variables while Pearson's Chisquare test or Fischer's exact test were used for categorical variables. Multivariable Cox regression analysis was performed using likelihood ratio tests to find associations between survival outcomes and select demographic/clinical factors. The significance level of all statistical two-sided tests was defined a priori as <0.05. All analyses were performed using R Studio [27]. Of 769 patients enrolled, 118 (15.3%) never started DEC-C during the data collection period. The vast majority (79.7%) of these patients were discharged from the program for reasons outlined in Figure 1, while others (n = 24, 20.3%) were pending treatment initiation. Of the 651 patients who initiated treatment, 427 patients discontinued treatment during the study period with the most common reasons being death (38.4%), disease progression (33.1%), and physician decision (13.3%). Two hundred and seven patients (26.9%) were still on treatment at the data collection cut-off date ( Figure 1).

Across
Of 769 patients enrolled, 118 (15.3%) never started DEC-C during the data collect period. The vast majority (79.7%) of these patients were discharged from the program reasons outlined in Figure 1, while others (n = 24, 20.3%) were pending treatment ini tion. Of the 651 patients who initiated treatment, 427 patients discontinued treatment d ing the study period with the most common reasons being death (38.4%), disease progr sion (33.1%), and physician decision (13.3%). Two hundred and seven patients (26.9 were still on treatment at the data collection cut-off date ( Figure 1). Stratifying patients by IPSS risk group, there were 389 patients with HR-MDS a 277 with LR-MDS. While there were more patients who died in the higher-risk subgro (n = 97, 44.3%) versus the lower-risk subgroup (n = 49, 32.0%), this is expected, and difference was not statistically significant. There were no statistically significant diff ences in treatment duration or red blood cell/platelet transfusion dependence statu enrollment between the HR-MDS and LR-MDS subgroups ( Table 2). Of 155 patients w received the 6th cycle re-enrollment form, data from 120 (77%) patients were collected these patients, 108 patients reported both IPSS score and the number of dose reducti in the last six cycles. Fifty-eight patients (53.7%) did not require any dose reduction. Co paring IPSS risk groups, a greater proportion of patients with HR-MDS did not requ any dose reduction in the last six cycles compared to those with LR-MDS (60.3% vs. 42. p = 0.03). Eighty-eight patients had both IPSS status and red blood cell transfusion sta reported at the 6th cycle. The greatest proportion of patients had maintained red blo Stratifying patients by IPSS risk group, there were 389 patients with HR-MDS and 277 with LR-MDS. While there were more patients who died in the higher-risk subgroup (n = 97, 44.3%) versus the lower-risk subgroup (n = 49, 32.0%), this is expected, and the difference was not statistically significant. There were no statistically significant differences in treatment duration or red blood cell/platelet transfusion dependence status at enrollment between the HR-MDS and LR-MDS subgroups ( Table 2). Of 155 patients who received the 6th cycle re-enrollment form, data from 120 (77%) patients were collected. Of these patients, 108 patients reported both IPSS score and the number of dose reductions in the last six cycles. Fifty-eight patients (53.7%) did not require any dose reduction. Comparing IPSS risk groups, a greater proportion of patients with HR-MDS did not require any dose reduction in the last six cycles compared to those with LR-MDS (60.3% vs. 42.5%, p = 0.03). Eighty-eight patients had both IPSS status and red blood cell transfusion status reported at the 6th cycle. The greatest proportion of patients had maintained red blood cell transfusion independence since enrollment. Of 90 patients with both IPSS scores and platelet transfusion status reported at the 6th cycle, the greatest proportion had maintained platelet transfusion independence since enrollment. Both antibiotic prophylaxis status and IPSS status was reported in 108 patients during the re-enrollment process among which only eight patients (7.8%) received antibiotic prophylaxis in the last 6 cycles ( Table 2). Subgroup analyses based on treatment duration revealed that a significantly greater proportion of patients who received <4 cycles stopped treatment due to death, patient decision, or side effect compared to those who received ≥4 cycles (p < 0.001). Conversely, treatment discontinuation due to disease progression, physician decision, or therapy switch was more common in patients who received ≥ 4 cycles. No significant differences between treatment duration subgroups were shown for age, province, or time to treatment initiation (Table 3). At the time of the data cut-off, the median OS among all patients who received DEC-C was 21.6 months (95% confidence interval (CI) 18.5-inf). The median PFS overall was 10.7 months (95% CI 9.2-13.3). Differences in OS between IPSS risk groups were statistically significant (p = 0.03), while differences in PFS were not statistically significant (p = 0.45). Survival estimates are further summarized in Figures 2 and 3.     Multivariable cox regression analysis revealed no significant associations between OS, PFS and various demographic and clinical factors (i.e., patient age at enrollment, patient province, IPSS risk group, time to treatment initiation, blast count status, antibiotic prophylaxis, dose reductions, and dose delays) ( Table 4). Multivariable cox regression analysis revealed no significant associations between OS, PFS and various demographic and clinical factors (i.e., patient age at enrollment, patient province, IPSS risk group, time to treatment initiation, blast count status, antibiotic prophylaxis, dose reductions, and dose delays) ( Table 4).

Discussion
This study examines real-world oral DEC-C use for the treatment of MDS and CMML through the Taiho Pharma Canada PSP. To our knowledge, the use of DEC-C in the Canadian real-world setting has yet to be described in the literature since its approval by Health Canada. There were 769 patients who enrolled into the PSP during our study period, which represents a notable uptake of this novel oral therapeutic. Interest and preference toward oral formulations over IV/SC chemotherapy is well documented in the literature. Indeed, a recent 2022 study analyzing the online survey data from MDS patients revealed both a preference and perceived personal benefit from improved quality of life from receiving oral DEC-C in comparison to IV/SC treatment [21]. The distribution of IPSS risk subgroups within patients with MDS in our cohort was similar to that of representative patient samples in the existing literature, where the greatest proportion of patients was INT-1 followed by INT-2 and then high risk [14,26]. The median age of our cohort was 76 years old (Range: . This is similar to the Canadian median age of diagnosis of 75 years for both MDS and CMML [4,5], and it is consistent with epidemiological studies characterizing both MDS and CMML to be diseases of older age [28,29]. The median age of our cohort was approximately five years older than that of previous clinical trials exploring oral DEC-C for MDS and CMML [24][25][26]. This difference may be explained by the more extensive eligibility criteria that trials use, which builds study cohorts that may not represent the diverse population of patients for whom DEC-C may be prescribed in the real world. The median OS of patients who received oral DEC-C was 21.6 months. This was significantly less than the median OS in the phase 3 study ASCERTAIN which demonstrated a median OS of 31.7 months [30]. A large proportion of patients in our study stopped treatment due to death (n = 164, 38.4%) or disease progression (n = 133, 33.1%). In comparison, the ASCERTAIN phase 3 study reported treatment discontinuation for disease progression in only six patients (4.5%) while reporting no deaths while on treatment. These discrepancies may be explained by ASCERTAIN's more stringent enrollment criteria which included eligibility for IV decitabine administration, life expectancy of three or more months, adequate organ function, ECOG performance status of 0-1, and fewer than 1 prior cycle of azacitidine or decitabine [31]. As such, patients enrolled in the phase 3 ASCERTAIN study were likely healthier at baseline and at lower risk of earlier death or disease progression compared to the present study cohort, which comprised all for whom DEC-C was indicated. Additionally, the data collection period reflects a unique timeframe in society where the COVID-19 pandemic brought unprecedented challenges to both the individual patient and healthcare systems at large. MDS patients, by the nature of their mean age and disease, were more susceptible to downstream effects of the virus. From treatment delays, scarcity of transfusion products, to death from COVID-19 virus itself, MDS patients have been well described in the literature to have suffered multiple negative consequences [32,33]. Future studies may reveal the true extent of COVID-19 on survival outcomes data during this time.
Our study revealed a median treatment duration of 4.2 cycles (range 0.0-22.7) for patients who received therapy and similar durations between LR-and HR-MDS. This median duration is in keeping with one other study exploring real-world data of DEC-C in the United States [34]. However, the median treatment duration for patients in ASCERTAIN was significantly longer at nine cycles [30]. Research has demonstrated that the longer treatment duration of HMA is associated with improved clinical outcomes, including higher reported response rates, reduced AML transformation, and improved OS [35][36][37][38][39][40]. Specifically, clinical trial evidence has suggested that patients require four to six cycles of HMA therapy to achieve a clinical response [36,[41][42][43]. However, prior real-world data suggest many patients do not persist with HMA treatment after initiation, receiving less than four or six cycles or having gaps over 90 days between cycles [41]. This is consistent with Canadian data from the national MDS Registry, with one study finding 33% of patients with HR-MDS receive <4 cycles of subcutaneous hypomethylation therapy with short overall survival [44]. Delayed response to hypomethylation therapy is not unusual; hence, for patients with stable disease or those who achieve clinical response from HMA, the continuation of therapy remains closely associated with improved survival [45]. This is reflected in current guidelines from the National Comprehensive Cancer Network (NCCN) and the European Society for Medical Oncology (ESMO) [14,15]. Consistent with the evidence that longer treatment duration is associated with superior survival outcomes, our study found that a significantly greater proportion of patients who received <4 cycles stopped treatment due to death versus those who received ≥4 cycles (49.3% vs. 28.4%, p < 0.001).
Reasons for early HMA discontinuation can be clinical in nature (e.g., patient mortality, disease progression, adverse effects) and/or non-clinical (e.g., logistical challenges, provider inexperience, socioeconomic barriers). In this study, oral DEC-C is shown to be a well-tolerated drug with 5.2% of patients (n = 22) stopping therapy due to side effects/tolerability issues. DEC-C has a tolerability profile consistent with IV decitabine [46] but is not impacted by the complex logistics inherent with IV HMA treatment which may result in early discontinuation or dose delay [47]. Of the 119 patients in this study who were followed up at 6 cycles, 55 patients (46.2%) reported no dose delays in their therapy, 30 patients (25.2%) received 1 dose delay, 14 patients (11.8%) received 2 dose delays, while 20 patients (16.8%) received 3 or more dose delays. Although most patients experienced one dose delay or less, reasons for dose delay may have been of a clinical or non-clinical nature and will require further investigation.
Transfusion dependence is not uncommon among MDS and CMML patients [12]. Approximately 50-90% of patients will require red blood cell transfusions with 30-50% requiring ≥1 platelet transfusion [48]. Most patients (60.2%) in our study were red blood cell transfusion dependent at enrollment with some (16.1%) who were platelet transfusion dependent. Transfusion dependence is associated with poorer quality of life and survival outcomes in MDS. As HMAs have been shown to promote transfusion independence in both LR-MDS [49] and HR-MDS [50], reporting any change in transfusion status is valuable in assessing the real-world performance of DEC-C. Of patients receiving ≥6 cycles of DEC-C, 19.8% achieved a change in transfusion status from being red blood cell transfusion dependent at enrollment to independent at the 6th cycle versus 9.3% who shifted from red blood cell transfusion independent to dependent. Red blood cell transfusion independence was maintained in 40.7% of patients after 6 cycles.
The cytopenic complications of MDS and CMML and the associated predisposition to infections warrant the consideration of antibiotic prophylaxis. Current guidelines recommend prophylactic antibiotics use for patients at high risk of febrile neutropenia, profound protracted neutropenia, and patients on hematopoietic stem cell therapy [51]. Additional recommendations include anti-viral and anti-fungal coverage in specific clinical scenarios. In this study, 57.0% of patients were reported as maintaining baseline blast counts from enrollment to cycle 6. This is important to quantify, as hypomethylating agents may cause transient neutropenia especially during initial cycles of therapy and increasing risks of infection [35,37,52,53]. Multivariable cox regression models revealed no statistically significant impact of blast count on OS or PFS. During DEC-C treatment, 8.7% reported being on prophylactic antibiotics.
This study contains inherent limitations which should be considered. Due to the nature of data collection and necessary privacy, we were not permitted to link other data sources; thus, we lacked access to data points not captured within the PSP. These may include prior treatment history, leukemic transformation prior to therapy, and specifics on antibiotic prophylaxis. Another limitation was the inability to identify specific reasons for patients' decisions not to receive DEC-C nor the rationales for physician decisions to stop DEC-C. Parsing these decisions in future studies may allow for more thorough assessments of non-persistence due to logistical factors, patient preferences, or provider inexperience. Furthermore, while the PSP was effective in facilitating early access to oral DEC-C to Canadians, this dataset may not be fully representative in the setting of wider access involving public drug funding plans and private reimbursement. Finally, due to the retrospective nature of our study analysis, the associations explored in our analysis are susceptible to unmeasured confounding variables.
Our study highlights the role that PSP and compassionate use programs serve in providing access and financial support to Canadians who require novel therapeutics like HMAs. Most patients in this study (64.4%, n = 496) were reimbursed through Taiho Pharma Canada's compassionate support program. By enabling early patient access to oral DEC-C in the context of MDS and CMML, the program facilitates foundational research which may allow for more comprehensive decision making regarding public drug funding. Additionally, despite data supporting the use of HMAs in the treatment of HR-MDS and CMML, several studies have suggested that approximately half of patients with HR-MDS do not receive HMA therapy at all [41,[53][54][55][56][57]. As real-world data on oral DEC-C's performance in the management of MDS and CMML are limited, the PSP enables timely research to help address this knowledge gap and optimize HMA use. Data Availability Statement: Data will not be shared due to patient confidentiality, according to ethics approval for this study.