Summary
Despite therapeutic advances and improved survival, the long‐term economic burden of chronic myeloid leukaemia (CML) remains under‐recognised given evolving treatment practices and fluctuating drug costs. We aimed to estimate and project total prevalence costs for CML in Sweden from a healthcare sector perspective, representing the direct healthcare expenditures for all patients living with CML, based on real‐world drug and procedure prices. We used data from the Swedish Cancer Register and the Swedish CML register to estimate and project prevalence and costs. The estimated numbers of prevalent cases were 1808 (95% confidence interval [CI], 1604–2011) in 2025 and 2120 (95% CI, 1916–2325) in 2030, driven by stable incidence and improved survival. Despite increasing prevalence, the annual total direct healthcare costs for all prevalent CML patients in Sweden decreased from USD 40.04 million (95% CI, 33.70–46.40) in 2015 to USD 26.04 million (95% CI, 23.13–28.97) in 2025, then projected to slightly increase to USD 30.67 million (95% CI, 27.70–33.64) in 2030. While CML prevalence proportions are expected to increase, declining treatment costs may mitigate the burden on the Swedish healthcare system. These population‐based projections can inform long‐term planning and pricing strategies for CML care.
Keywords: Chronic myeloid, cost of illness, economic burden, leukaemia, prevalence, Sweden
INTRODUCTION
The introduction of tyrosine kinase inhibitors (TKIs) has transformed the treatment of chronic myeloid leukaemia (CML), turning a once fatal disease into a mostly manageable chronic condition. Survival among CML patients has improved substantially and now approaches that of the general population 1 ; meanwhile, the incidence of CML has remained relatively stable over recent decades. 2 , 3 The combination of improved survival and a steady incidence has led to a growing prevalence in Sweden, from 11.9 cases per 100 000 individuals in 2012 to a predicted 15.1 per 100 000 in 2020. 2
The economic burden of CML in Sweden is evolving, mainly due to declining drug prices linked to novel generic TKIs and emerging new therapies. Initially, TKIs were associated with high costs (more than € 30 000 per patient per year in Sweden 4 ), but those costs have been considerably reduced due to patent expiration. Generic imatinib emerged in the Swedish market in December 2016, 5 followed by the first generic dasatinib, which entered the market during June 2020. 6 Further price reductions are expected for nilotinib and bosutinib, corresponding to their patent expirations in Sweden in 2025 and in the European Union (EU) in 2028 7 , respectively. A novel third‐generation TKI, asciminib has recently been approved in Sweden 8 in a third‐line setting with promising molecular response rates and encouraging tolerability, but at higher costs.
Treatment‐free remission (TFR) has become an important goal for CML management, based on growing evidence that many patients who achieve a lasting deep molecular response (DMR) can safely discontinue therapy. 9 , 10 , 11 , 12 , 13 , 14 Clinical trials such as European Stop Tyrosine Kinase Inhibitor Trial (EURO‐SKI) 9 and population‐based studies in Sweden 13 suggest that 40%–50% of eligible patients with DMR can enter TFR, although nearly half may have molecular relapse after TFR within 24 months and thus need to re‐initiate TKI treatment. Nevertheless, the possibility of temporary or sustained TFR offers a new clinical management with potential for cost savings. For example, EURO‐SKI estimated € 22 million in drug cost savings among the study population of 586 patients. 9
The interplay of incidence, survival, prevalence, drug prices and evolving treatment strategies complicates projections of the future economic burden of CML. A few studies 15 , 16 , 17 have examined the long‐term healthcare costs associated with CML. However, none has used nationwide, whole‐population cancer registers that fully account for temporal changes in treatment practices, drug pricing and changes in patient composition across different disease states. This study aims to address this gap by estimating the total prevalence costs of CML in Sweden from 2015 to 2030, using data from the Swedish Cancer Register and the Swedish CML register.
METHODS
Study design and data sources
We used several nationwide Swedish registers, including the Swedish Cancer Register and the Swedish CML register, to obtain data on survival, demographics, disease characteristics and treatment patterns for patients diagnosed with CML between ages 18 and 99 years. For cost data, we used individual‐level data from the Swedish CML register and the National Prescribed Drug Register, as well as aggregate‐level data from several official price lists and studies. Table 1 provides a summary of the data sources used in the analysis. Detailed definitions of variables and data extraction procedures are presented in Section S1.
TABLE 1.
Overview of data sources.
| Data type | Data source | Period |
|---|---|---|
| Survival (CML patients) | Swedish Cancer Register, Cause of Death Register | Diagnosed 1973–2019 with follow‐up until 2020 |
| Mortality (general population) | Human Mortality Database 18 | Up to 2023 (projected constant thereafter) |
| Population size (general population) | Human Mortality Database 18 | Up to 2024 (projected constant thereafter) |
| Disease phase at diagnosis (CML patients) | Swedish CML register | Diagnosed 2002–2023 |
| TKI treatment patterns | Swedish CML register, National Prescribed Drug Register | Diagnosed 2007–2017 with follow‐up until 2018 |
| TFR | Swedish CML register | Diagnosed 2007–2012 with follow‐up until 2018 |
| Disease progression to AP/BP | Swedish CML Register, Swedish Cancer Register | Diagnosed 2007–2017 with follow‐up until 2018 |
| AlloSCT procedures | Swedish CML register, Swedish Patient Register | Diagnosed 2007–2017 with follow‐up until 2018 |
| Costs for TKI (individual‐level) | Swedish CML register, National Prescribed Drug Register | Diagnosed 2007–2017 with follow‐up until 2018 |
| Costs for TKI (aggregate‐level) | TLV Price and Decision Database 19 | 2019–2025 (projected constant thereafter) |
| Costs for alloSCT and progression | Hernlund et al., 20 adjusted yearly by the Extra‐regional Pricing List for EU/EEA and Switzerland 21 | Diagnosed 2007–2015 |
| Costs for laboratory testing | National Price List for Laboratory Medical Services 22 | 2024 |
Note: Details of variable definitions of each data type are provided in Section S1.
Abbreviations: AlloSCT, allogeneic stem cell transplantation; AP, accelerated phase; BP, blastic phase; CML, chronic myeloid leukaemia; TFR, treatment‐free remission; TKI, tyrosine kinase inhibitor; TLV, the Dental and Pharmaceutical Benefits Agency (Tandvårds‐ och läkemedelsförmånsverket).
Estimating and projecting prevalence and total prevalence costs
We used the Prevalence, Incidence, Analysis MODel method 23 to estimate and project prevalence, incorporating incidence rates and net survival (described below) together with the population and mortality data of Sweden.
We estimated total prevalence costs from a healthcare sector perspective. 24 Total prevalence costs refer to the total healthcare expenditures incurred for all individuals living with CML. For each year, the costs were obtained by multiplying the average yearly cost per patient in each disease state by the number of patients in that state and summing across all states. Full statistical details and the overview workflow are provided in Section S2.
Estimating and projecting of incidence
Because prevalence estimates depend on both incidence and survival, we first modelled CML incidence using an age–period–cohort (APC) model fitted to observed data from the Swedish Cancer Register between 1973 and 2019. The final model was then used to project CML incidence rates from 2020 to 2030. Details of the APC modelling approach are provided in Section S3.
Estimating and projecting of net survival
Net survival was estimated from individual‐level data on CML patients diagnosed between 1973 and 2019, representing the probability of surviving CML in the absence of other causes of death. 25 To reflect different future survival assumptions, we considered two scenarios for projecting prevalence, both within a relative survival framework 26 :
Scenario 1—Fixed survival (conservative projection): We estimated period‐specific in 3‐year intervals net survival for patients diagnosed during 1973–1975, 1975–1977, …, and 2015–2019 using the Pohar–Perme estimator. 27 We assumed no improvements in net survival beyond 2019, fixed at the 2019 level for projections from 2020 to 2030.
Scenario 2—Dynamic survival (optimistic projection): We fitted a flexible parametric relative survival model 28 to the observed data up to 2019 and projected survival from 2020 to 2030 using the estimated model coefficients. This approach allowed for continued improvements over time by extrapolating the temporal trends.
Scenario 1 served as the base case for subsequent analyses, with Scenario 2 results reported in sensitivity analyses. Full statistical methods for both scenarios are presented in Section S4.
Estimating the distribution of prevalent CML patients across disease states
Patients with CML can be classified into various disease states. 29 To estimate the distribution of patients across these states over time, we used both observed (2000–2019) and projected (2020–2030) incidence data. We estimated the distribution of patients across seven disease states over time—including first‐line TKI (1L TKI), second‐line TKI (2L TKI), third‐line or later TKI (3L+ TKI), TFR, progression, allogeneic stem cell transplantation (alloSCT) and death using the CML natural history model previously developed by Chen et al. 29 Methodological details for this model are described elsewhere. 29
Finally, we multiplied the estimated proportion of each state by the total number of prevalent cases to determine the number of patients in each disease state from 2015 to 2030. Details of the estimation procedures are provided in Section S5 and Table S1.
Estimating average yearly cost per patient
TKI costs were estimated using data from the Swedish CML register, the National Prescribed Drug Register and the Price and Decision Database from the Dental and Pharmaceutical Benefits Agency (Tandvårds‐ och läkemedelsförmånsverket, TLV). Costs for progression, alloSCT and other health resources were informed by the published literature 4 , 20 and Sweden's official price lists. 21 , 22 All costs were inflated to 2024 Swedish krona (SEK) using the consumer price index from Statistics Sweden 30 and converted to USD using the 2024 exchange rate (1 USD = 10.56 SEK) from Sveriges Riksbank (the central bank of Sweden). 31 Detailed cost calculations for the average yearly cost per patient are provided in Section S6 and Tables S2–S4.
We did not apply discounting to future costs, as our study aims to examine trends in economic burden rather than evaluate the cost‐effectiveness of specific interventions. However, for readers who would like to apply our results to assess specific interventions, we have provided the results with 3% and 5% discounting rates to costs from 2025 onwards in Tables S5 and S6 respectively.
Sensitivity analysis
We performed one‐way sensitivity analyses to evaluate the influence of various parameters on the projections. The parameters considered were as follows: (1) Scenario 2, representing dynamic survival; (2) variations in discount rates on costs: 3% and 5% applied from 2025 onwards (Tables S5 and S6), if a decision maker wishes to evaluate any intervention using our projections; (3) a 25% increase in yearly cost of 3L+ TKI in 2024, reflecting the market introduction of asciminib as a third‐line TKI in 2023 8 ; (4) a 25% decrease in yearly 3L+ TKI costs in 2025, consequent to the potential entry of generic bosutinib in 2024 7 ; (5) a ±25% variation in yearly 1L TKI costs in 2029; (6) a ±25% variation in yearly 2L TKI costs in 2029; and (7) a ±25% variation in yearly 3L+ TKI costs in 2029, reflecting potential price reductions following the potential introduction of generic nilotinib and ponatinib 7 respectively.
We developed an interactive R Shiny application (R version 4.4.2) to facilitate sensitivity analyses on prevalent cases, yearly cost per patient and total prevalence costs. The app integrates editable tables, plots and with options to apply discounting rate, allowing users to reproduce and extend our analyses. It is publicly accessible at: https://enochytchen.shinyapps.io/CMLEcoBurdenSE/.
Role of the funding source and ethical approval
This work was supported by Vetenskapsrådet (Swedish Research Council) and Cancerfonden (Swedish Cancer Society). The funding sources had no role in the study design, data collection, data analysis or writing of the manuscript. The study was approved by the Swedish Ethical Review Authority (DNR 2020‐05425, 2020‐06544 and 2021‐02472).
RESULTS
The observed and predicted crude incidence rates for CML among patients aged 55–64 years at diagnosis from 1973 to 2030, stratified by sex, are shown in Figure 1. In general, the predicted incidence rates for males and females followed the same trend. Although male incidence was higher than female in the 1970s, our model predicted a higher incidence rate in females, 2.0 per 100 000 person‐years (95% confidence intervals [CI]: 1.8–2.2), than in males, 1.8 per 100 000 person‐years (95% CI: 1.6–2.0), in 2030. We presented the crude incidence rates of all age groups in Figure S2.
FIGURE 1.
Observed and predicted crude incidence rates (per 100 000 person‐years) for male and female patients with chronic myeloid leukaemia aged 55–64 years at diagnosis in Sweden during 1973–2030. Black dots represent observed incidence rates. Red lines represent the estimates by the age–period–cohort model, with shaded areas indicate 95% confidence intervals. Dashed lines after 2019 represent projections beyond the observed period.
Net survival after a CML diagnosis has dramatically improved over recent decades. Figure 2 displays survival trends for patients aged 55–64 years at diagnosis under two different scenarios, shown here as a representative example. In Scenario 1 (fixed survival), net survival rose steadily from the mid‐1970s to early 2000s; by 2015–2017, 1‐year net survival exceeded 95% and 5‐year survival reached roughly 80%–90%, with similar patterns in both sexes. For patients diagnosed after 2019, survival was assumed to remain at the level estimated for the 2015–2019 cohort. Net survival estimates for all age groups are provided in Figure S3 (Scenario 1: fixed survival) and Figure S4 (Scenario 2: dynamic survival).
FIGURE 2.
Observed and projected net survival for male and female patients with chronic myeloid leukaemia aged 55–64 years at diagnosis in Sweden during 1973–2030. Black and red points represent observed 1‐ and 5‐year net survival estimates by the Pohar–Perme estimator respectively (Scenario 1: fixed survival). For patients diagnosed after 2019, survival was assumed to remain at the level estimated for the 2015–2019 cohort. Black and red lines represent 1‐ and 5‐year net survival estimates by the flexible parametric relative survival model, respectively, with 95% confidence intervals (Scenario 2: dynamic survival).
Projections of CML incidence and prevalence under Scenario 1 (fixed survival) and Scenario 2 (dynamic survival) are provided in Table 2 and illustrated by prevalence curves in Figure S5. Overall, the estimates in prevalence were similar but slightly higher in Scenario 2 than in Scenario 1. Similarly, the increase in the predicted number of prevalent cases from 2015 to 2030 was greater in Scenario 2 (82.9%, 95% CI: 72.5%–97.2%) than in Scenario 1 (75.9%, 95% CI: 66.5%–88.9%), which reflects the change in prevalence over that period.
TABLE 2.
Population, estimated incidence and prevalence of chronic myeloid leukaemia aged 18–99 years at diagnosis in Sweden under two scenarios: (1) fixed and (2) dynamic survival, in 2015, 2020, 2025 and 2030, presented with crude and age‐standardised (to the 2000 Swedish population) estimates with 95% confidence interval.
| Scenario | Year | Population in thousands | Incident cases | Incidence per 100 000 person‐years | Age‐standardised incidence per 100 000 person‐years | Prevalent cases | Prevalence per 100 000 persons | Age‐standardised prevalence per 100 000 persons |
|---|---|---|---|---|---|---|---|---|
| Scenario 1. Fixed survival | 2015 | 9745 | 103.3 (99.6–107.0) | 1.06 (0.98–1.14) | 1.02 (0.94–1.10) | 1205.4 (1014.3–1396.5) | 12.36 (10.40–14.33) | 11.84 (9.88–13.80) |
| 2020 | 10 255 | 113.2 (108.4–117.9) | 1.11 (1.01–1.20) | 1.06 (0.96–1.15) | 1497.4 (1296.4–1698.4) | 14.60 (12.64–16.56) | 13.87 (11.91–15.83) | |
| 2025 | 10 363 | 121.9 (117.1–126.7) | 1.18 (1.08–1.27) | 1.10 (1.00–1.20) | 1807.5 (1604.3–2010.6) | 17.44 (15.48–19.40) | 16.05 (14.09–18.01) | |
| 2030 | 10 420 | 130.4 (125.7–135.2) | 1.25 (1.15–1.35) | 1.16 (1.06–1.26) | 2120.3 (1916.1–2324.6) | 20.35 (18.39–22.31) | 18.24 (16.28–20.20) | |
| Scenario 2. Dynamic survival | 2015 | 9745 | 103.3 (99.6–107.0) | 1.06 (0.98–1.14) | 1.02 (0.94–1.10) | 1209.3 (1018.2–1400.4) | 12.41 (10.45–14.37) | 11.92 (9.96–13.89) |
| 2020 | 10 255 | 113.2 (108.4–117.9) | 1.11 (1.01–1.20) | 1.06 (0.96–1.15) | 1543.0 (1342.0–1744.0) | 15.05 (13.09–17.01) | 14.30 (12.34–16.26) | |
| 2025 | 10 363 | 121.9 (117.1–126.7) | 1.18 (1.08–1.27) | 1.10 (1.00–1.20) | 1881.8 (1678.7–2084.9) | 18.15 (16.19–20.11) | 16.65 (14.69–18.61) | |
| 2030 | 10 420 | 130.4 (125.7–135.2) | 1.25 (1.15–1.35) | 1.16 (1.06–1.26) | 2211.9 (2007.7–2416.2) | 21.23 (19.27–23.19) | 18.90 (16.94–20.86) |
From 2015 to 2030, CML in Sweden shows an increasing number of prevalent cases but declining average yearly costs, resulting in reduced total prevalence costs. These trends are illustrated in Figure 3 and summarised in Table 3: total prevalence costs decreased from USD 40.04 million (95% CI 33.70–46.40) in 2015 to USD 26.04 million (95% CI 23.13–28.97) in 2025, and then slightly increased to USD 30.67 million (95% CI 27.70–33.64) in 2030, as the average yearly cost per patient decreased from USD 33 224 in 2015 to USD 14 408 in 2025 and USD 14 464 in 2030. Patient numbers are expected to grow mainly in 1L and 2L TKI (Table 3). First‐year alloSCT is expected to be the costliest state (more than USD 150 000 per patient yearly), while the progression cost is at USD 20 000–30 000. TKI costs will decline over time: average 1L costs drop from USD 35 244 in 2015 to USD 8063 in 2025, and 2L/3L+ costs fall from roughly USD 40 000 to USD 20 000. TFR is consistently predicted to be the least expensive state (USD 7558 in 2015 and USD 6030 in 2025).
FIGURE 3.
Estimated and projected prevalent cases, average yearly cost per patient (USD), total prevalence costs (USD thousand) of chronic myeloid leukaemia in Sweden from 2015 to 2030.
TABLE 3.
Prevalent cases, yearly cost per patient and prevalence costs by state and overall total.
| Year | 1L TKI | 2L TKI | 3L TKI | Progression | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | |
| 2015 | 523.2 (440.3, 606.2) | 35 244 | 18.44 (15.52, 21.36) | 284.0 (239.0, 329.0) | 40 467 | 11.49 (9.67, 13.31) | 162.3 (136.6, 188.1) | 39 520 | 6.41 (5.40, 7.43) | 74.3 (62.5, 86.1) | 19 269 | 1.43 (1.20, 1.66) |
| 2016 | 542.5 (459.7, 625.4) | 34 245 | 18.58 (15.74, 21.42) | 293.7 (248.8, 338.5) | 40 264 | 11.83 (10.02, 13.63) | 175.3 (148.5, 202.1) | 36 216 | 6.35 (5.38, 7.32) | 78.8 (66.7, 90.8) | 23 465 | 1.85 (1.57, 2.13) |
| 2017 | 547.2 (466.2, 628.2) | 13 311 | 7.28 (6.21, 8.36) | 311.8 (265.7, 358.0) | 35 929 | 11.20 (9.55, 12.86) | 192.4 (163.9, 220.9) | 28 093 | 5.41 (4.60, 6.21) | 81.7 (69.6, 93.8) | 23 557 | 1.92 (1.64, 2.21) |
| 2018 | 565.6 (484.5, 646.7) | 12 114 | 6.85 (5.87, 7.83) | 323.9 (277.5, 370.3) | 36 851 | 11.94 (10.23, 13.65) | 208.7 (178.8, 238.6) | 34 393 | 7.18 (6.15, 8.21) | 79.2 (67.9, 90.6) | 24 845 | 1.97 (1.69, 2.25) |
| 2019 | 564.5 (486.0, 643.0) | 9201 | 5.19 (4.47, 5.92) | 338.9 (291.8, 386.0) | 31 735 | 10.75 (9.26, 12.25) | 226.8 (195.3, 258.3) | 28 623 | 6.49 (5.59, 7.39) | 88.4 (76.1, 100.7) | 25 630 | 2.27 (1.95, 2.58) |
| 2020 | 579.2 (501.5, 657.0) | 12 108 | 7.01 (6.07, 7.95) | 344.4 (298.2, 390.6) | 32 246 | 11.11 (9.62, 12.60) | 241.4 (209.0, 273.8) | 32 641 | 7.88 (6.82, 8.94) | 94.6 (81.9, 107.2) | 26 718 | 2.53 (2.19, 2.86) |
| 2021 | 592.3 (515.3, 669.2) | 10 727 | 6.35 (5.53, 7.18) | 352.5 (306.7, 398.2) | 24 141 | 8.51 (7.40, 9.61) | 255.5 (222.3, 288.7) | 27 897 | 7.13 (6.20, 8.05) | 99.7 (86.8, 112.7) | 26 628 | 2.65 (2.31, 3.00) |
| 2022 | 607.6 (531.4, 683.8) | 11 228 | 6.82 (5.97, 7.68) | 362.8 (317.3, 408.4) | 23 899 | 8.67 (7.58, 9.76) | 270.4 (236.5, 304.4) | 28 010 | 7.57 (6.62, 8.53) | 103.3 (90.3, 116.2) | 25 229 | 2.61 (2.28, 2.93) |
| 2023 | 627.2 (551.4, 703.0) | 9147 | 5.74 (5.04, 6.43) | 375.2 (329.9, 420.6) | 19 665 | 7.38 (6.49, 8.27) | 286.4 (251.8, 321.0) | 23 365 | 6.69 (5.88, 7.50) | 104.7 (92.1, 117.4) | 25 264 | 2.65 (2.33, 2.97) |
| 2024 | 641.6 (566.9, 716.3) | 10 943 | 7.02 (6.20, 7.84) | 388.7 (343.5, 434.0) | 19 847 | 7.71 (6.82, 8.61) | 302.6 (267.4, 337.9) | 25 048 | 7.58 (6.70, 8.46) | 111.6 (98.6, 124.6) | 26 409 | 2.95 (2.60, 3.29) |
| 2025 | 656.4 (582.7, 730.2) | 8063 | 5.29 (4.70, 5.89) | 401.4 (356.3, 446.5) | 17 929 | 7.20 (6.39, 8.01) | 318.9 (283.0, 354.7) | 20 304 | 6.47 (5.75, 7.20) | 117.8 (104.6, 131.1) | 21 088 | 2.48 (2.21, 2.76) |
| 2026 | 671.0 (598.1, 743.9) | 8063 | 5.41 (4.82, 6.00) | 413.5 (368.6, 458.4) | 17 929 | 7.41 (6.61, 8.22) | 334.7 (298.4, 371.1) | 20 304 | 6.80 (6.06, 7.53) | 124.0 (110.5, 137.4) | 21 088 | 2.61 (2.33, 2.90) |
| 2027 | 685.2 (613.1, 757.3) | 8063 | 5.52 (4.94, 6.11) | 425.3 (380.5, 470.0) | 17 929 | 7.63 (6.82, 8.43) | 350.3 (313.4, 387.1) | 20 304 | 7.11 (6.36, 7.86) | 130.1 (116.4, 143.8) | 21 088 | 2.74 (2.45, 3.03) |
| 2028 | 698.9 (627.6, 770.2) | 8063 | 5.64 (5.06, 6.21) | 436.9 (392.3, 481.4) | 17 929 | 7.83 (7.03, 8.63) | 365.4 (328.1, 402.7) | 20 304 | 7.42 (6.66, 8.18) | 136.3 (122.4, 150.2) | 21 088 | 2.87 (2.58, 3.17) |
| 2029 | 712.2 (641.6, 782.8) | 8063 | 5.74 (5.17, 6.31) | 448.2 (403.8, 492.6) | 17 929 | 8.04 (7.24, 8.83) | 380.1 (342.4, 417.8) | 20 304 | 7.72 (6.95, 8.48) | 142.4 (128.3, 156.6) | 21 088 | 3.00 (2.71, 3.30) |
| 2030 | 725.1 (655.3, 795.0) | 8063 | 5.85 (5.28, 6.41) | 459.3 (415.0, 503.5) | 17 929 | 8.23 (7.44, 9.03) | 394.4 (356.4, 432.4) | 20 304 | 8.01 (7.24, 8.78) | 148.6 (134.3, 162.9) | 21 088 | 3.13 (2.83, 3.44) |
| Year | AlloSCT, first year | AlloSCT, second year | TFR | Total | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | Prevalent cases | YCPP (USD) | Prevalence costs (USD m) | Prevalent cases | Average YCPP (USD) | Prevalence costs (USD m) | |
| 2015 | 6.0 (5.0, 7.0) | 159 432 | 0.96 (0.80, 1.12) | 71.2 (59.9, 82.5) | 9448 | 0.67 (0.57, 0.78) | 84.3 (71.0, 97.7) | 7558 | 0.64 (0.54, 0.74) | 1205.4 (1014.3, 1396.5) | 33 224 | 40.04 (33.70, 46.40) |
| 2016 | 11.0 (9.3, 12.7) | 161 312 | 1.77 (1.50, 2.05) | 73.3 (62.1, 84.5) | 9559 | 0.70 (0.59, 0.81) | 89.6 (75.9, 103.3) | 7529 | 0.67 (0.57, 0.78) | 1264.2 (1071.2, 1457.3) | 33 026 | 41.75 (35.37, 48.14) |
| 2017 | 6.0 (5.1, 6.9) | 163 525 | 0.98 (0.83, 1.13) | 87.7 (74.7, 100.7) | 9690 | 0.85 (0.72, 0.98) | 96.8 (82.5, 111.1) | 7442 | 0.72 (0.61, 0.83) | 1323.6 (1127.8, 1519.5) | 21 432 | 28.36 (24.16, 32.58) |
| 2018 | 8.0 (6.9, 9.1) | 165 354 | 1.32 (1.14, 1.50) | 94.9 (81.3, 108.5) | 9799 | 0.93 (0.80, 1.06) | 103.3 (88.5, 118.1) | 6994 | 0.72 (0.62, 0.83) | 1383.6 (1185.2, 1581.9) | 22 339 | 30.91 (26.50, 35.33) |
| 2019 | 8.0 (6.9, 9.1) | 177 076 | 1.42 (1.22, 1.61) | 105.2 (90.5, 119.8) | 10 493 | 1.10 (0.95, 1.26) | 110.6 (95.2, 125.9) | 7004 | 0.77 (0.67, 0.88) | 1442.4 (1241.9, 1642.9) | 19 413 | 27.99 (24.11, 31.89) |
| 2020 | 8.0 (6.9, 9.1) | 181 050 | 1.45 (1.25, 1.65) | 113.8 (98.6, 129.1) | 10 729 | 1.22 (1.06, 1.39) | 116.0 (100.4, 131.6) | 7008 | 0.81 (0.70, 0.92) | 1497.4 (1296.4, 1698.4) | 21 376 | 32.01 (27.71, 36.31) |
| 2021 | 8.0 (7.0, 9.0) | 183 544 | 1.47 (1.28, 1.65) | 122.6 (106.7, 138.5) | 10 877 | 1.33 (1.16, 1.51) | 121.0 (105.3, 136.7) | 6906 | 0.84 (0.73, 0.94) | 1551.6 (1350.1, 1753.0) | 18 228 | 28.28 (24.61, 31.94) |
| 2022 | 8.0 (7.0, 9.0) | 219 014 | 1.75 (1.53, 1.97) | 132.1 (115.5, 148.6) | 12 979 | 1.71 (1.50, 1.93) | 125.9 (110.1, 141.6) | 6468 | 0.81 (0.71, 0.92) | 1610.1 (1408.1, 1812.0) | 18 604 | 29.94 (26.19, 33.72) |
| 2023 | 8.0 (7.0, 9.0) | 209 838 | 1.68 (1.47, 1.89) | 142.4 (125.2, 159.6) | 12 435 | 1.77 (1.56, 1.98) | 130.9 (115.0, 146.7) | 6037 | 0.79 (0.69, 0.89) | 1674.8 (1472.4, 1877.2) | 15 937 | 26.70 (23.46, 29.93) |
| 2024 | 8.0 (7.1, 8.9) | 216 422 | 1.73 (1.54, 1.93) | 153.2 (135.4, 171.1) | 12 825 | 1.96 (1.74, 2.19) | 135.8 (120.0, 151.6) | 6054 | 0.82 (0.73, 0.92) | 1741.6 (1538.8, 1944.4) | 17 101 | 29.77 (26.33, 33.24) |
| 2025 | 8.0 (7.1, 8.9) | 211 175 | 1.69 (1.50, 1.88) | 164.3 (145.9, 182.8) | 12 514 | 2.06 (1.83, 2.29) | 140.7 (124.8, 156.5) | 6030 | 0.85 (0.75, 0.94) | 1807.5 (1604.3, 2010.6) | 14 408 | 26.04 (23.13, 28.97) |
| 2026 | 8.0 (7.1, 8.9) | 211 175 | 1.69 (1.50, 1.88) | 175.6 (156.5, 194.7) | 12 514 | 2.20 (1.96, 2.44) | 145.5 (129.7, 161.3) | 6030 | 0.88 (0.78, 0.97) | 1872.3 (1668.9, 2075.7) | 14 420 | 27.00 (24.06, 29.94) |
| 2027 | 8.0 (7.2, 8.8) | 211 175 | 1.69 (1.52, 1.86) | 187.0 (167.3, 206.6) | 12 514 | 2.34 (2.09, 2.59) | 150.3 (134.4, 166.1) | 6030 | 0.91 (0.81, 1.00) | 1936.1 (1732.4, 2139.8) | 14 431 | 27.94 (24.99, 30.88) |
| 2028 | 8.0 (7.2, 8.8) | 211 175 | 1.69 (1.52, 1.86) | 198.4 (178.2, 218.7) | 12 514 | 2.48 (2.23, 2.74) | 154.9 (139.1, 170.7) | 6030 | 0.93 (0.84, 1.03) | 1998.7 (1794.8, 2202.6) | 14 443 | 28.86 (25.92, 31.82) |
| 2029 | 8.0 (7.2, 8.8) | 211 175 | 1.69 (1.52, 1.86) | 209.9 (189.1, 230.7) | 12 514 | 2.63 (2.37, 2.89) | 159.3 (143.5, 175.0) | 6030 | 0.96 (0.87, 1.06) | 2060.1 (1856.0, 2264.2) | 14 453 | 29.78 (26.83, 32.73) |
| 2030 | 8.0 (7.2, 8.8) | 211 175 | 1.69 (1.52, 1.86) | 221.5 (200.1, 242.8) | 12 514 | 2.77 (2.50, 3.04) | 163.5 (147.7, 179.2) | 6030 | 0.99 (0.89, 1.08) | 2120.3 (1916.1, 2324.6) | 14 464 | 30.67 (27.70, 33.64) |
Note: 95% confidence intervals are presented in parentheses.
Abbreviations: 1L, first‐line; 2L, second‐line; 3L+, third‐line and later; AlloSCT, allogeneic stem cell transplantation; TKI, tyrosine kinase inhibitor; USD m, USD millions; YCPP, yearly cost per patient.
Results from the one‐way sensitivity analyses for total prevalence costs in 2025 and 2030 are presented in Figure 4. Using dynamic survival (Scenario 2) had only a modest effect on total prevalence costs, with increases of USD 0.43 million in 2025 and USD 0.57 million in 2030. In contrast, changes in 3L+ TKI pricing had a more notable impact: a 25% cost increase in 2024 led to an increase of USD 1.62 million in 2025, while a 25% price decrease in 2025 reduced costs by USD 1.62 million. The most substantial change was observed in 2030, where applying a discounting rate of 5% from 2025 and onwards reduced total prevalence costs by USD 7.78 million. Adjusting 1L, 2L and 3L+ TKI costs by ±25% from 2029 resulted in cost differences of USD ±1.46, ±2.06, and ±2.00 million respectively. An interactive R Shiny app is available to reproduce and further explore the sensitivity analyses.
FIGURE 4.
One‐way sensitivity analysis of total prevalence costs of chronic myeloid leukaemia in Sweden for the years 2025 and 2030. Bars represent the difference from the base case (USD million) under various scenario assumptions. Black bars indicate lower costs, and red bars indicate higher costs relative to the base case.
DISCUSSION
In this study, we projected a steady increase in the number of individuals living with CML in Sweden from 2015 to 2030 under both fixed and dynamic survival scenarios, despite stable incidence. The increase was predominantly driven by the rising numbers of patients on 1L and 2L TKIs. Notably, while the prevalent cases increased over the years, the total prevalence costs decreased from USD 40.04 million (95% CI 33.70–46.40) in 2015 to USD 30.67 million (95% CI 27.70–33.64) in 2030. Despite a growing number of patients, the total prevalence costs remained relatively stable mainly because the average yearly cost of TKI treatment declined over time.
Our projections highlight the uncertainty introduced by our survival assumptions. The fixed survival scenario (Scenario 1) may be overly conservative, as net survival, particularly in younger patients, would likely improve over time (Figure S3). Conversely, the dynamic survival scenario (Scenario 2) may be overly optimistic for older age groups; for instance, it predicted 5‐ and 10‐year net survival exceeding 90% for patients aged 75–99 years after 2020 (Figure S4), a pattern likely reflecting model extrapolation rather than clinical reality.
The declining cost trends observed in our analysis were primarily driven by changes in TKI pricing. The introduction of generic imatinib 5 in December 2016 and generic dasatinib 6 in June 2020 substantially reduced 1L to 3L+ costs in Sweden. Future potential generic entry of nilotinib and bosutinib, following expected EU patent expiry in 2028, 7 may further reduce TKI costs. Assuming these generics reach price levels comparable to today's generic imatinib (approximately 800 USD per patient per year), their yearly cost per patient could decline by an estimated 60%–80% relative to 2025 prices. While TFR contributes to cost savings, its overall impact on total prevalence costs is modest, as approximately half of patients initiating TFR resume treatment within 24 months. 13 Likewise, while first‐year alloSCT remains highly expensive, only a small number of patients fall within this state annually, 3 resulting in a limited effect on total prevalence costs. The same applies to patients in progression (accelerated phase/blastic phase), who also represent a small proportion of the CML population.
Our findings broadly align with and extend previous studies. Gunnarsson et al. 2 predicted 15.1 CML cases per 100 000 persons in 2020 and 18.2 in 2030 under a fixed survival scenario, whereas our model estimated 14.60 and 20.35 cases for 2020 and 2030, respectively. These differences likely reflect our use of more recent follow‐up data and different modelling approaches: Gunnarsson applied a period analysis and assumed constant excess mortality after 10 years, while we used empirical relative survival estimates without such constraints. Similarly, a recent French study 32 projected 18.74 prevalent cases per 100 000 persons in 2020 and 24.24 in 2030, which is in line with the rising trend observed in our estimates in Sweden.
Cost of illness studies, including ours, necessitate periodic updates as clinical practice and prices change. In our cost modelling, we incorporated disease states from a natural history model 23 informed by the Swedish CML register, with pricing data from Sweden's Prescribed Drug Register and TLV's Price and Decision Database. 19 We also treated TFR as a distinct disease state to capture its potential cost‐saving impact. As clinical practice, treatment patterns and prices evolve, these estimates should be updated.
This study has several strengths. To our knowledge, it is the first nationwide, whole‐population cost study for projecting CML prevalence costs using cancer register data, including both the Swedish Cancer Register and the Swedish CML register. 33 These results are specific to Sweden, although the methods are adaptable to countries with similarly structured population registers and comparable treatment patterns. We employed a previously developed natural history model for CML treatments 29 to predict the distribution of patients across disease states and integrated individual‐level pricing data from the National Prescribed Drug Register and pharmaceutical and pricing register and database. 19 Projecting costs under two survival scenarios (fixed vs. dynamic survival) allowed us to quantify future uncertainty and showed that alternative survival assumptions had only a modest impact on the predicted number of prevalent cases and consequently the prevalence costs of CML in Sweden.
Nonetheless, several limitations warrant consideration. First, pricing assumptions for the years 2019–2025 relied on price ratios relative to 2018, and we assumed constant prices beyond 2025. Although discussions with experts in haematology suggested that substantial price shifts beyond 2025 are unlikely, this assumption might lead to a slight overestimation of costs in the later projection years, as historical trends indicated a decline in prices. Second, while the composition of TKI treatment was based on empirical data from 2007 to 2018 (Figure S6), we assumed that the distribution at each treatment line would remain constant beyond 2018. This may not fully capture future changes in treatment guidelines and prescribing patterns. Third, although adverse events, such as cardiovascular complications associated with nilotinib use, 34 and hospitalisation costs are part of direct health expenditures, these components were not included in our estimates. Notably, hospitalisation alone was reported to account for approximately 8.6% of the total yearly cost of care for CML patients, 35 suggesting that our estimates may modestly underestimate the true economic burden. Nonetheless, the one‐way sensitivity analyses provided a structured approach to explore key sources of uncertainty in our projections. Fourth, our TFR data were limited to patients diagnosed between 2007 and 2012; extrapolation of these cohorts to later years may overestimate or underestimate the true number of patients eligible for TFR. Finally, cost estimates for patients in progression and alloSCT states were based on the aggregate data from Hernlund et al. 20 and adjusted yearly by Sweden's official price list, 21 rather than individual‐level data from registers. Nevertheless, because relatively few patients transition to progression or alloSCT, the impact of these estimates on total prevalence costs is likely to be minimal.
Future policy initiative should (1) optimise prescribing to favour cost‐effective TKIs when clinically appropriate, (2) expand eligibility for TFR when safety permits and (3) maintain post‐marketing surveillance of newer generations of TKIs, such as asciminib, to confirm their long‐term value for healthcare budget.
In conclusion, our study offers 15‐year estimations and projections of total prevalence costs for CML that inform healthcare budget planning in Sweden, which are ready inputs for future economic evaluations and solid evidence for clinicians and policymakers aiming to manage CML care efficiently.
AUTHOR CONTRIBUTIONS
P.W.D. and T.D. acquired and cleaned the data. E.Y.‐T.C. and T.D. cleaned the data. E.Y.‐T.C. and F.D.M. conducted the formal statistical analyses. E.Y.‐T.C., P.W.D., F.D.M., M.S.C., and S.H. developed the statistical methodology. E.Y.‐T.C. and S.H. conceptualized the project and drafted the manuscript. E.Y.‐T.C. prepared all figures and visualizations. E.Y.‐T.C., P.W.D., F.D.M., T.D., L.S., M.B., M.S.C., and S.H. critically reviewed and revised the manuscript.
CONFLICT OF INTEREST STATEMENT
Enoch Yi‐Tung Chen, Paul W. Dickman, Fabrizio Di Mari, Takeda. Mark S. Clements and Shuang Hao: No relationship to disclose. Torsten Dahlén: (all outside of the submitted work) Consulting: Xspray pharma, Novartis. Leif Stenke: (all outside of the submitted work) Consulting: Xspray Pharma. Magnus Björkholm: (all outside of the submitted work) Grant committee: Incyte; Educational programme committee: Roche, Pfizer, Bristol Myers Squibb, Abbvie, Janssen, Takeda, Novartis; Consulting: WntResearch, Janssen‐Cilag, and Schain Research; Research grant.
Supporting information
Data S1.
ACKNOWLEDGEMENTS
The study was supported by grants from Cancerfonden (the Swedish Cancer Society) and Vetenskapsrådet (the Swedish Research Council). The authors thank Xiaoyang Du and Daniela Skalt for their comments on the RShiny application.
Chen EY‐T, Dickman PW, Di Mari F, Dahlén T, Stenke L, Björkholm M, et al. Empirical and projected economic burden of chronic myeloid leukaemia in Sweden from 2015 to 2030: A population‐based study. Br J Haematol. 2025;207(6):2441–2450. 10.1111/bjh.70193
DATA AVAILABILITY STATEMENT
The statistical codes that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to ethical restrictions.
REFERENCES
- 1. Bower H, Björkholm M, Dickman PW, Höglund M, Lambert PC, Andersson TML. Life expectancy of patients with chronic myeloid leukemia approaches the life expectancy of the general population. JCO. 2016;34(24):2851–2857. [DOI] [PubMed] [Google Scholar]
- 2. Gunnarsson N, Sandin F, Höglund M, Stenke L, Björkholm M, Lambe M, et al. Population‐based assessment of chronic myeloid leukemia in Sweden: striking increase in survival and prevalence. Eur J Haematol. 2016;97(4):387–392. [DOI] [PubMed] [Google Scholar]
- 3. Dahlén T, Olsson K, Sandin F. Chronic myeloid leukemia (CML) ‐ national quality report for 2023 (In Swedish: Kronisk myeloisk leukemi (KML) ‐ Nationell kvalitetsrapport för 2023). 2024.
- 4. Ohm L, Lundqvist A, Dickman P, Höglund M, Persson U, Stenke L, et al. Real‐world cost‐effectiveness in chronic myeloid leukemia: the price of success during four decades of development from non‐targeted treatment to imatinib. Leuk Lymphoma. 2015;56(5):1385–1391. [DOI] [PubMed] [Google Scholar]
- 5. The Dental and Pharmaceutical Benefits Agency (In Swedish: Tandvårds‐ och läkemedelsförmånsverket, TLV) . The development of pharmaceutical expenditure in Sweden – managed entry agreements is an increasingly important tool for cost control as well as for early and equal access. Report No.: 3610/2017. 2017.
- 6. The Dental and Pharmaceutical Benefits Agency (In Swedish: Tandvårds‐ och läkemedelsförmånsverket, TLV) . Dasatinib Sandoz is included in the high‐cost coverage with limitation (Swedish: Dasatinib Sandoz ingår i högkostnadsskyddet med begränsning). 2020.
- 7. Venkatesan S, Lamfers M, Leenstra S, Vulto AG. Overview of the patent expiry of (non‐)tyrosine kinase inhibitors approved for clinical use in the EU and the US. GaBI J. 2017;6(2):89–96. [Google Scholar]
- 8. The Dental and Pharmaceutical Benefits Agency (In Swedish: Tandvårds‐ och läkemedelsförmånsverket, TLV) . Underlag för beslut om subvention ‐ Nyansökan Nämnden för läkemedelsförmåner‐‐Scemblix (asciminib) [Internet]. 2023. Available from: https://www.tlv.se/download/18.187f44431872c3a93f232608/1680255944998/bes230323_scemblix_3266‐2022_underlag.pdf
- 9. Saussele S, Richter J, Guilhot J, Gruber FX, Hjorth‐Hansen H, Almeida A, et al. Discontinuation of tyrosine kinase inhibitor therapy in chronic myeloid leukaemia (EURO‐SKI): a prespecified interim analysis of a prospective, multicentre, non‐randomised, trial. Lancet Oncol. 2018;19(6):747–757. [DOI] [PubMed] [Google Scholar]
- 10. Etienne G, Guilhot J, Rea D, Rigal‐Huguet F, Nicolini F, Charbonnier A, et al. Long‐term follow‐up of the French stop imatinib (STIM1) study in patients with chronic myeloid leukemia. J Clin Oncol. 2017;35(3):298–305. [DOI] [PubMed] [Google Scholar]
- 11. Kimura S, Imagawa J, Murai K, Hino M, Kitawaki T, Okada M, et al. Treatment‐free remission after first‐line dasatinib discontinuation in patients with chronic myeloid leukaemia (first‐line DADI trial): a single‐arm, multicentre, phase 2 trial. Lancet Haematol. 2020;7(3):e218–e225. [DOI] [PubMed] [Google Scholar]
- 12. Ross DM, Masszi T, Gómez Casares MT, Hellmann A, Stentoft J, Conneally E, et al. Durable treatment‐free remission in patients with chronic myeloid leukemia in chronic phase following frontline nilotinib: 96‐week update of the ENESTfreedom study. J Cancer Res Clin Oncol. 2018;144(5):945–954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Flygt H, Sandin F, Dahlén T, Dremaine A, Lübking A, Markevärn B, et al. Successful tyrosine kinase inhibitor discontinuation outside clinical trials—data from the population‐based Swedish chronic myeloid leukaemia registry. Br J Haematol. 2021;193(5):915–921. [DOI] [PubMed] [Google Scholar]
- 14. Hochhaus A, Baccarani M, Silver RT, Schiffer C, Apperley JF, Cervantes F, et al. European LeukemiaNet 2020 recommendations for treating chronic myeloid leukemia. Leukemia. 2020;34(4):966–984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lin PJ, Winn AN, Parsons SK, Neumann PJ, Weiss ES, Cohen JT. Linking costs and survival in the treatment of older adults with chronic myeloid leukemia: an analysis of SEER‐Medicare data from 1995 to 2007. Med Care. 2016;54(4):380–385. [DOI] [PubMed] [Google Scholar]
- 16. Kenzik KM, Bhatia R, Williams GR, Bhatia S. Medicare and patient spending among beneficiaries diagnosed with chronic myelogenous leukemia. Cancer. 2019;125(15):2570–2578. [DOI] [PubMed] [Google Scholar]
- 17. Rivera DR, Enewold L, Barrett MJ, Banegas MP, Filipski KK, Freedman AN, et al. Population‐based utilization and costs associated with tyrosine kinase inhibitors for first‐line treatment of chronic myelogenous leukemia among elderly patients. JMCP. 2020;26(12):1494–1504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Max Planck Institute for Demographic Research (Germany) , University of California, Berkeley (USA) , French Institute for Demographic Studies (France) . Human mortality database [Internet]. Available from: www.mortality.org
- 19. The Dental and Pharmaceutical Benefits Agency (In Swedish: Tandvårds‐ och läkemedelsförmånsverket, TLV) . TLV's price and decision database (In Swedish: TLV:s pris‐ och beslutsdatabas) [Internet]. Available from: https://www.tlv.se/ovriga‐sidor/sok‐i‐databasen.html
- 20. Hernlund E, Redig J, Paulsson B, Rangert Derolf Å, Höglund M, Vertuani S, et al. Socioeconomic cost of AML in Sweden—a population‐based study using multiple nation‐wide registers. eJHaem. 2021;2(3):385–393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Karolinska University Hospital (In Swedish: Karolinska Universitetssjukhuset) . Extra‐regional pricing list for EU/EEA and Switzerland (In Swedish: Utomregional prislista samt för EU/EES och Schweiz), Karolinska Universitetssjukhuset. 2025.
- 22. Karolinska University Hospital (In Swedish: Karolinska Universitetssjukhuset) . National price list for laboratory medical services (In Swedish: Nationell Prislista för Laboratoriemedicinsk Service). 2024.
- 23. Verdecchia A, Capocaccia R, Egidi V, Golini A. A method for the estimation of chronic disease morbidity and trends from mortality data. Stat Med. 1989;8(2):201–216. [DOI] [PubMed] [Google Scholar]
- 24. Sanders GD, Neumann PJ, Basu A, Brock DW, Feeny D, Krahn M, et al. Recommendations for conduct, methodological practices, and reporting of cost‐effectiveness analyses: second panel on cost‐effectiveness in health and medicine. JAMA. 2016;316(10):1093. [DOI] [PubMed] [Google Scholar]
- 25. Ederer F, Axtell LM, Cutler SJ. The relative survival rate: a statistical methodology. Natl Cancer Inst Monogr. 1961;6:101–121. [PubMed] [Google Scholar]
- 26. Dickman PW, Coviello E. Estimating and modeling relative survival. Stata J. 2015;15(1):186–215. [Google Scholar]
- 27. Perme MP, Stare J, Estève J. On estimation in relative survival. Biometrics. 2012;68(1):113–120. [DOI] [PubMed] [Google Scholar]
- 28. Nelson CP, Lambert PC, Squire IB, Jones DR. Flexible parametric models for relative survival, with application in coronary heart disease. Stat Med. 2007;26(30):5486–5498. [DOI] [PubMed] [Google Scholar]
- 29. Chen EY, Dahlén T, Stenke L, Björkholm M, Hao S, Dickman PW, et al. Loss in overall and quality‐adjusted life expectancy for patients with chronic‐phase chronic myeloid leukemia. Eur J Haematol. 2025;114(2):334–342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Statistics Sweden (In Swedish: Statistikmyndigheten) . Consumer price index, fixed index numbers (1980=100). [cited 2025 March 28] [Internet]. Available from: https://www.scb.se/en/finding‐statistics/statistics‐by‐subject‐area/prices‐and‐economic‐trends/price‐statistics/consumer‐price‐index‐cpi/pong/tables‐and‐graphs/consumer‐price‐index‐cpi/cpi‐fixed‐index‐numbers‐1980100/
- 31. Sweden Central Bank (In Swedish: Riksbanken) . Annual average exchange rates. [cited 2025 March 28] [Internet]. Available from: https://www.riksbank.se/en‐gb/statistics/search‐interest‐‐exchange‐rates/annual‐average‐exchange‐rates/
- 32. Delord M, Foulon S, Cayuela JM, Rousselot P, Bonastre J. The rising prevalence of chronic myeloid leukemia in France. Leuk Res. 2018;69:94–99. [DOI] [PubMed] [Google Scholar]
- 33. Höglund M, Sandin F, Hellström K, Björeman M, Björkholm M, Brune M, et al. Tyrosine kinase inhibitor usage, treatment outcome, and prognostic scores in CML: report from the population‐based Swedish CML registry. Blood. 2013;122(7):1284–1292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Dahlén T, Edgren G, Lambe M, Höglund M, Björkholm M, Sandin F, et al. Cardiovascular events associated with use of tyrosine kinase inhibitors in chronic myeloid leukemia: a population‐based cohort study. Ann Intern Med. 2016;165(3):161. [DOI] [PubMed] [Google Scholar]
- 35. Breccia M, Chiodi F, Nardozza AP, Valsecchi D, Perrone V, Sangiorgi D, et al. The economic burden of chronic myeloid leukemia in patients with later lines: findings from a real‐world analysis in Italy. Adv Ther. 2023;40(3):961–974. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1.
Data Availability Statement
The statistical codes that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to ethical restrictions.