An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Azacitidine (Onureg): CADTH Reimbursement Review: Therapeutic area: Acute myeloid leukemia [Internet]. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2022 Jan.
Submitted for Review.
Summary of Economic Evaluation.
Evidence from the CADTH Clinical Review and the QUAZAR AML-001 trial suggested that oral azacitidine led to a statistically significant improvement in relapse-free survival (RFS) and overall survival (OS) among patients with acute myeloid leukemia (AML) who were in remission after chemotherapy when compared to best supportive care (BSC).
CADTH attempted to address limitations identified in the sponsor’s base case by considering dose extension in the calculation of oral azacitidine costs in a revised base case. Both the sponsor and CADTH base cases suggest that oral azacitidine is not cost-effective compared to BSC at conventionally accepted willingness-to-pay thresholds. The CADTH base case suggests that the incremental cost-effectiveness ratio (ICER) for oral azacitidine plus BSC is $355,554 per quality-adjusted life-year (QALY) compared with BSC alone. A price reduction of at least 85% would be required to reach a willingness-to-pay threshold of $50,000 per QALY.
The cost-effectiveness of oral azacitidine was highly sensitive to statistical models used to fit RFS and OS data. Other key drivers included health utility values and the treatment duration. Findings from the CADTH reanalyses likely underestimate the true value of the ICER and the price reduction, as CADTH could not address the critical limitation regarding the implied post-relapse survival benefit of oral azacitidine due to constraints introduced by the submitted model structure. Oral azacitidine is not likely to be cost-effective at conventional thresholds without a large reduction in price.
This section is a summary of the feedback received from the patient groups, registered clinicians, and drug plans that participated in the CADTH review process (specifically, information that pertains to the economic submission).
One patient group, the Leukemia and Lymphoma Society of Canada, provided input for the review of oral azacitidine for the treatment of AML via an online survey created by the society. Patients stated that a range of AML symptoms negatively affects their physical, psychological, and social well-being. Patients reported that fatigue, nausea, numbness or body aches, being immunocompromised, losing vision in an eye, among other symptoms, contributed to being unable to participate in daily life, work, or exercise. Most patients noted that they received chemotherapy and a stem cell or bone marrow transplant after diagnosis, while others had received drug or radiation therapy and 1 patient had received chimeric antigen receptor T-cell therapy. Altogether, patients reported poor experiences with available AML treatments, particularly with chemotherapy and stem cell transplants, and expressed that various treatment-related side effects on their physical health, mental health, and social and educational development had negative impacts on their quality of life. None of the patients reported having experience with oral azacitidine; however, patients expressed hope that any new AML drug or treatment would help maintain remission, result in fewer side effects, be offered at a lower cost, and be accessible in their neighbourhoods. Additional social and emotional support was also desired by patients. Last, patients noted the negative impacts of AML on their caregivers, such as stress, worry, and sadness.
Registered clinician input described an unmet need for AML patients previously treated with induction therapy, with or without consolidation therapy, who are in complete remission (CR) or complete remission with incomplete blood count recovery (CRi) and ineligible for a transplant. Registered clinicians stated that there are no other maintenance treatments available for the highly lethal cancer with poor outcomes, as patients are typically followed by surveillance thereafter. Registered clinicians further noted that maintenance azacitidine has been the first treatment to demonstrate an OS benefit in the decades since treatment with 7 days of cytarabine plus 3 days of daunorubicin became the standard of care. Registered clinicians indicated that oral azacitidine may be continued in the absence of an overt clinical relapse and may be discontinued upon disease progression or treatment intolerance. The clinician group further highlighted the ease of use of oral azacitidine as it can be administered at home.
The drug plans highlighted considerations for the implementation of oral azacitidine that are relevant to the economic analysis. Drug plan input noted several factors that may affect patient treatment costs. As the list price per tablet is high, dose adjustments or modifications and days of extended treatment may have a significant impact on treatment costs and contribute to a higher risk of drug wastage, even if the drug were to be supplied in 7-day blister packs. Additionally, drug plans noted costs associated with granulocyte colony-stimulating factor, which is required to treat febrile neutropenia, a possible adverse effect of oral azacitidine.
Several of these concerns were addressed in the sponsor’s model:
CADTH addressed some of these concerns as follows:
The current review is for oral azacitidine (Onureg) for adult patients with AML who achieved CR or CRi following induction therapy with or without consolidation treatment and who are not eligible for hematopoietic stem cell transplant (HSCT).
The sponsor submitted a cost-utility analysis comparing costs and outcomes for oral azacitidine in combination with BSC (oral azacitidine) with no active therapy combined with BSC (no active therapy) for adults with AML who achieved CR or CRi following induction therapy with or without consolidation treatment, and who are not eligible for HSCT. The modelled population was in line with the reimbursement request and Health Canada–approved indication.
Oral azacitidine is available as 200 mg and 300 mg tablets. The recommended dosage is 300 mg once daily for 14 days every 28-day treatment cycle. The per-cycle cost of oral azacitidine 300 mg was estimated to be $19,992. The sponsor focused on oral azacitidine 300 mg and the 14-day dosing schedule in its base case. No drug costs were assigned to the comparator of no active therapy. The sponsor assumed 100% compliance in its calculations.
The clinical outcome was QALYs and life-years. The economic analysis was undertaken over a lifetime horizon of 20 years from the perspective of a Canadian publicly funded health care system. Costs and QALYs were discounted at a rate of 1.5% per annum.
The sponsor submitted a partitioned survival model (PSM) with 3 health states: RFS, relapse, and death (Appendix 3, Figure 1). The proportion of patients who were relapse-free, who experienced relapse, or who were dead at any time over the model horizon was derived using joint curves and the log-normal distribution. All patients entered the RFS and were assumed to receive treatments until more than 15% of blasts were observed in peripheral blood or bone marrow or until unacceptable toxicity occurred. Patients could discontinue treatment but remain in the RFS health state. At the end of each cycle, the proportion of patients who relapsed or died was derived based on the area under the survival curves. Specifically, OS was partitioned to estimate the proportion of patients in the death state, while OS and RFS curves were used to estimate the proportion of patients in the relapse-free health state. The difference between the OS and RFS curves was partitioned at each time point to estimate the proportion of patients in the relapse health state. Relapse was defined as the earliest date of any of the following: 5% or more bone marrow blasts from the central pathology report; the appearance of greater than 0% blasts in the peripheral blood with a later bone marrow confirmation (bone marrow blast ≥ 5%) within 100 days; or at least 2 peripheral blasts of 5% or more within 30 days.
The modelled population reflected the baseline patient characteristics of the QUAZAR AML-001 trial,1 an ongoing phase III, international, multicentre, randomized, double-blind, placebo-controlled study comparing the efficacy and safety of oral azacitidine plus BSC versus placebo plus BSC as maintenance therapy among patients with AML who are in CR or CRi after intensive chemotherapy. The submitted model assumed a mean age of 67.9 years, a mean weight of ||||| kg, and an average height of |||| cm.
The RFS and OS curves for oral azacitidine and no active therapy were generated using patient-level data from the QUAZAR AML-001 trial (data cut-off: July 15, 2019). The sponsor predicted OS and RFS curves using a joint curve model, which fitted a log-normal model to individual patient–level data using a treatment indicator (azacitidine versus no active therapy) as a covariate. The log-normal distribution was selected for both RFS and OS data based on clinical validity and statistical fit.
The model accounted for grade 3 or 4 AEs occurring in at least 5% of patients in the treatment arm of the QUAZAR AML-001 trial population. The proportion of patients receiving a subsequent therapy, and the mix of subsequent therapies, was informed by the QUAZAR AML-001 trial and was validated by clinical experts. The sponsor modelled HSCT as part of subsequent therapy and assumed that the proportions of receiving HSCT were 6.3% for patients receiving oral azacitidine and 13.7% for patients on no active therapy, as reported in the QUAZAR AML-001 trial.
State-specific health utility values were based on a UK study2 that used a time trade-off methodology to elicit health state utilities for AML from 210 members of the UK general population. The sponsor did not use health utility values measured in the QUAZAR AML-001 trial because the trial did not capture data on health-related quality of life for patients beyond the treatment period and into relapse. Utility decrements were included in the model to capture the impact of AEs and HSCT on health-related quality of life.
Costs included those associated with drugs (acquisition and premedication), disease management, subsequent therapy, AEs, and end-of-life care. The cost of oral azacitidine was based on sponsor data on file and a mean time on treatment of |||| cycles reported in the QUAZAR AML-001 trial. Subsequent treatment costs were estimated based on the proportion of patients receiving each subsequent therapy, per-cycle drug acquisition, treatment administration costs obtained from the published sources, and the estimated treatment duration. The model also considered disease management costs for physician and nurse visits, laboratory tests, chemistry and liver panels, blood transfusions, and bone marrow aspirates and biopsies. The frequency of resource use for each health state and treatment arm was informed by clinical experts. The proportion of patients receiving red blood cell and platelet transfusions in relapse was informed by the QUAZAR AML-001 trial. Costs of BSC were included in the model to capture ongoing disease management costs. It was assumed that there were no differences in BSC between patients receiving oral azacitidine and those on no active therapy. Costs related to HSCT were based on the average cost per procedure in Ontario, as reported by Ontario Health. Unit costs for each AE were obtained from the Ontario Case Costing Initiative database and the 2020 Ontario Schedule of Benefits, Physician Services. The cost of end-of-life care was applied as a one-off cost to patients who transitioned to the “death” health state; the costs were based on a study by Bekelman et al. (2016),3 which reported hospital costs for the 30 days preceding death for cancer patients in Canada (Ontario) and 6 other developed countries.
All analyses were run probabilistically with 5,000 iterations, with the deterministic and probabilistic results being comparable. The probabilistic findings are presented here. No drug costs were assigned to patients in the group receiving no active therapy.4
In the sponsor’s base-case analysis, oral azacitidine was associated with an ICER of $348,988 per QALY compared to no active therapy over a lifetime horizon (Table 3). At a willingness-to-pay threshold of $50,000 per QALY, the probability of oral azacitidine being cost-effective was 0% compared to no active therapy.4
The main driver of the incremental costs was drug cost (94.1%), followed by HSCT costs (2.9%) and disease management costs (1.2%). At the end of the model time horizon of 20 years, 1.9% of patients in the model were still alive. A breakdown of the sponsor-submitted results for the base-case population by trial duration and extrapolation period shows that 38.2% of the expected QALY gains come from the time beyond the trial period.
Summary of Sponsor’s Economic Evaluation.
Additional results from the sponsor’s submitted economic evaluation base case are presented in Appendix 3.
The sponsor performed scenario analyses by varying discount rates, reducing the time horizon to 15 years, using a societal perspective, considering dose extension or reduction, using alternative health utility values, and using alternative parametric survival models to predict RFS and OS curves. Treatment efficacy was the most influential driver of the cost-effectiveness results. Compared to no active therapy, the estimated ICERs of oral azacitidine ranged between $247,532 (assuming the lower bound of hazard ratios for RFS and OS) and $1,090,868 (assuming the upper bound of hazard ratios for RFS and OS) per QALY. Other key determinants included time on treatment and health utility values.
CADTH identified several key limitations to the sponsor's analysis that have notable implications on the economic analysis:
Additional uncertainty around comparative effectiveness was contributed by the sponsor’s use of a joint curve model with log-normal distribution to predict RFS and OS data during the trial and after the trial. Upon visual inspection, the predicted RFS and OS data slightly overestimated the Kaplan-Meier (KM) curves observed in the QUAZAR AML-001 trial; this overestimate may reduce the ICER of oral azacitidine.
An additional limitation was identified but was not considered to be key:
Key assumptions made by the sponsor and appraised by CADTH are provided in Table 4.
Key Assumptions of the Submitted Economic Evaluation (Not Noted as Limitations to the Submission).
CADTH could not adequately address some limitations (i.e., inability to account for the impact of subsequent treatments, including HSCT, on post-relapse survival and OS) of the sponsor’s model due to the model structure. The CADTH base case was derived by considering dose extension in the calculation of oral azacitidine costs. Dose reduction was not included because Canadian pharmacies are likely to dispense the full quantity of oral medications for each treatment cycle. In addition, CADTH removed a management fee for oral chemotherapy and calculated the per-cycle cost of ondansetron based on 14 days of ondansetron 8 mg once daily.
Table 5 details the change made to derive the CADTH base case, and the summary results of the CADTH base case are presented in Table 6. Additional results are shown in Appendix 4 (Table 11). Results from the CADTH base case suggested that oral azacitidine was associated with higher costs ($322,869) and increased QALYs (0.91), for an ICER of $355,456 per QALY. The estimated ICER was slightly higher than that reported in the sponsor’s base case due to increased oral azacitidine costs when considering dose extension and the longer duration of ondansetron premedication. The probability that oral azacitidine is cost-effective was zero at the willingness to value of $50,000 per QALY. Approximately 39% of the QALY gains and 2% of incremental costs came from the extrapolation of outcomes beyond the trial follow-up period.
CADTH Revisions to the Submitted Economic Evaluation.
Summary of the Stepped Analysis of CADTH Reanalysis Results.
Based on the CADTH base case, scenario analyses were conducted. These analyses explored the impact of the following model parameters and assumptions: survival distributions to extrapolate RFS and OS data; health utility values; costs of end-of-life care; time on treatment duration; the proportion of patients receiving post-relapse HSCT; and a perspective of analysis.
Results from scenario analyses (Appendix 4, Table 11) demonstrated that the cost-effectiveness findings were sensitive to how RFS and OS data were incorporated into the economic model. In a scenario in which KM curves were used to estimate RFS and OS data during the trial period and parametric survival models were used to predict long-term RFS and OS data, the ICER increased from $355,456 to $543,359 per QALY. The ICERs were also influenced by health utility values and the treatment duration. Using alternative health utility values from Tremblay et al. (2020),6 which reported a lower utility value for the RFS state but a slightly higher utility value for the relapse state, increased the ICER to $384,118. As expected, assuming a similar proportion of patients receiving post-relapse HSCT for the groups receiving oral azacitidine and no active therapy and extending the duration of treatment for oral azacitidine increased the estimated ICERs. Cost-effectiveness findings were found to be robust to the changes in the costs of end-of-life care when considered from a societal perspective.
CADTH undertook a price reduction analysis based on the sponsor's base case and CADTH’s base case (Table 7). The results show that a price reduction of 85% is required for oral azacitidine to be considered cost-effective at a willingness-to-pay threshold of $50,000 per QALY. Given the uncertainty surrounding the existence of a post-relapse survival benefit, the price reduction is likely underestimated.
CADTH Price Reduction Analyses.
Evidence from the CADTH Clinical Review and the QUAZAR AML-001 trial results suggested that oral azacitidine led to significant improvement in RFS and OS among patients with AML who were in remission after chemotherapy. Survival models used to estimate the RFS and OS drove the incremental difference in QALYs, while the time on oral azacitidine was the key driver of the incremental costs between oral azacitidine and no active therapy.
CADTH identified several limitations within the sponsor’s submission: the lack of clinical evidence supporting post-relapse survival benefits of oral azacitidine shown in the economic analysis; the lack of fit of survival models used to estimate RFS and OS data; the exclusion of dose extension from the calculation of oral azacitidine costs; and the assumption on the rates of post-relapse HSCT. CADTH attempted to address these limitations by performing the following reanalyses: considering dose extension in the calculation of oral azacitidine costs, removing a management fee for oral chemotherapy, using an alternative regimen for ondansetron premedication, using alternate survival distributions to extrapolate RFS and OS data, applying alternate health utility values, assigning alternative end-of-life care costs, varying of time on treatment duration, changing the proportion of patients receiving post-relapse HSCT, and adopting a societal perspective. Both the sponsor’s and CADTH’s base cases suggested that oral azacitidine was associated with higher costs and improved QALYs but was not cost-effective compared to no active therapy. At the submitted price, the probability that oral azacitidine is cost-effective was 0%. A price reduction of at least 85% would be required to make oral azacitidine a cost-effective option at a willingness-to-pay threshold of $50,000 per QALY, although this is likely an underestimate.
The cost-effectiveness of oral azacitidine was highly sensitive to statistical approaches used to fit RFS and OS data. If the KM curves used to reflect RFS and OS data during the trial period and survival models were used to predict the long-term RFS and OS data, the ICER would increase substantially from $355,456 to $543,359 per QALY. Other key drivers included health utility values and the treatment duration. Using a lower utility value for the RFS state but a slightly higher utility value for the relapse state increased the ICER from $355,456 to $384,118 per QALY. Moreover, if the type of maintenance therapy in the first remission was assumed to have no impact on the proportion of patients receiving post-relapse HCST, the ICER would rise to $366,705 per QALY. Cost-effectiveness findings were robust to the changes in the cost of end-of-life care and the use of a societal perspective.
The CADTH estimates of cost-effectiveness are likely biased in favour of oral azacitidine, as the reanalysis was unable to address the critical limitation regarding the unproven post-relapse survival benefit due to constraints introduced by the submitted model structure. As such, oral azacitidine is unlikely to be cost-effective at a conventional threshold without a large reduction in price.
adverse event
acute myeloid leukemia
best supportive care
complete remission
complete remission with incomplete blood count recovery
hematopoietic stem cell transplant
incremental cost-effectiveness ratio
Kaplan-Meier
overall survival
partitioned survival model
quality-adjusted life-year
relapse-free survival
Note this appendix has not been copy-edited.
The comparators presented in Table 8 have been deemed to be appropriate based on feedback from clinical expert(s) and drug plans. Comparators may be recommended (appropriate) practice or actual practice. Existing product listing agreements are not reflected in the table and as such, the table may not represent the actual costs to public drug plans.
CADTH Cost Comparison Table for Acute Myeloid Leukemia.
Note this appendix has not been copy-edited.
Submission Quality.
Note this appendix has not been copy-edited.
Model Structure.
Observed and Predicted Overall Survival Data — Oral Azacitidine.
Observed and Predicted Overall Survival Data — No Active Therapy.
Observed and Predicted Relapse-Free Survival — Oral Azacitidine.
Observed and Predicted Relapse-Free Survival — No Active Therapy.
Note this appendix has not been copy-edited.
Disaggregated Summary of CADTH’s Economic Evaluation Results.
Note this appendix has not been copy-edited.
Summary of Key Take-Aways.
The sponsor assessed the budget impact of the introduction of oral azacitidine for patients with AML who achieve CR or CRi following induction therapy, with or without consolidation treatment, and who are not eligible for HSCT, from the drug plan perspective in the Canadian setting (excluding Quebec), over a 3-year time horizon.8 The sponsor only included drug acquisition costs. In the reference scenario, the sponsor assumed that patients would no active therapy (i.e., BSC), which was associated with no costs. In the new drug scenario, oral azacitidine was assumed to displace market shares of BSC.8
The sponsor estimated the eligible population size using an epidemiological approach, by leveraging data from the QUAZAR trial, multiple sources in the literature, and assumptions based on clinical expert input.8
Key inputs to the Budget Impact Analysis are documented in Table 13.
Sponsor’s Estimation of the Size of the Eligible Population.
Summary of Key Model Parameters.
Results of the sponsor’s base-case analysis reflected the drug program perspective and revealed that the introduction of oral azacitidine in patients with AML who achieved CR or CRi following induction therapy with or without consolidation treatment, and who are not eligible for hematopoietic stem cell transplant, would result in an incremental budget impact of $14,713,983 in year 1, $22,931,743 in year 2, $27,797,095 in year 3, for a total budget impact of $65,442,821 over the 3-year time horizon.8
CADTH identified several key limitations to the sponsor’s analysis that have notable implications on the results of the Budget Impact Analysis:
A table noting the changes made to the sponsor’s BIA as part of the CADTH reanalysis is available in Table 14.
CADTH Revisions to the Submitted Budget Impact Analysis.
Applying these changes in Table 17 resulted in an increase in the budget impact under the drug plan perspective to $100,647,777 over 3 years. The results of the CADTH stepwise reanalyses are presented in summary format in Table 15, and a more detailed breakdown is presented in Table 16.
Summary of the CADTH Reanalyses of the Budget Impact Analysis.
CADTH also conducted additional scenario analyses to address the remaining uncertainty regarding the potential size of the eligible population:
Detailed Breakdown of the CADTH Reanalyses of the Budget Impact Analysis.
Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND), a copy of which is available at http://creativecommons.org/licenses/by-nc-nd/4.0/
Your browsing activity is empty.
Activity recording is turned off.
See more...
