Global welfare burden of hematologic malignancies from 2015 to 2023.

Introduction

The global burden of disease is undergoing a profound demographic transition driven by declining fertility and sustained gains in life expectancy. By 2050, individuals aged 65 years and older are projected to comprise approximately 16% of the global population. 1 , 2 Beyond the increasing prevalence of chronic diseases, population aging alters the biological landscape of cancer risk by weakening immune surveillance through immunosenescence and age-associated clonal hematopoiesis, thereby increasing susceptibility to hematologic malignancies. 3 , 4 , 5 , 6 Consistent with these demographic and biological shifts, both the incidence and the associated economic burden of hematologic malignancies have risen substantially between 1990 and 2021, with marked heterogeneity across regions. 7 Unlike most solid tumors, hematologic malignancies exhibit a distinctive bimodal age distribution. Acute lymphoblastic leukemia (ALL) predominates in childhood, whereas acute myeloid leukemia (AML) and diffuse large B-cell lymphoma (DLBCL) primarily affect middle-aged and older adults. Consequently, hematologic malignancies generate welfare losses across the entire life course, including premature mortality in early life, erosion of human capital during working ages, and prolonged morbidity in later life, resulting in a uniquely persistent and multifaceted societal burden. 8 , 9

Over the past decade, the therapeutic landscape of hematologic malignancies has evolved rapidly with the introduction of targeted agents, immunotherapies, and cellular therapies. 10 , 11 , 12 These advances have translated into meaningful survival gains for selected patient populations but have also been accompanied by unprecedented escalation in treatment costs. Evidence from SEER-Medicare analyses and real-world studies indicates that the cumulative cost of managing AML across the induction, remission, and relapse phases can exceed several hundred thousand dollars. For example, venetoclax-based regimens are associated with annual costs of approximately USD $215,000 to $251,000. 13 The financial burden is further amplified by cellular therapies, such as chimeric antigen receptor T cell (CAR-T) products, with per-patient costs commonly ranging from $375,000 to $475,000, excluding downstream costs related to toxicity management and long-term supportive care. 14

To mitigate the fiscal impact of high-cost therapies, developed countries have increasingly adopted value-based pricing mechanisms and risk-sharing arrangements, such as managed entry agreements, to balance budgetary sustainability with access to innovation. In contrast, low- and middle-income countries frequently experience prolonged delays in the availability of affordable generics owing to secondary patent protections, regulatory barriers, and constrained market access, leaving patients exposed to persistently high prices for novel agents. 15 , 16 , 17 Consequently, as innovation premiums rise in tandem with population aging, traditional morbidity and mortality indicators are insufficient to capture the broader macroeconomic consequences of hematologic malignancies. This inadequacy is especially evident in middle-income countries, where growth in health insurance financing often lags behind GDP expansion, amplifying the affordability gap for high-cost cancer therapies. 16 , 18 , 19

Existing assessments of the economic burden of hematologic malignancies have largely concentrated on direct medical expenditures or productivity losses estimated through human capital approaches, with limited attention to the monetization of welfare losses associated with pain, disability, and premature mortality. 7 , 20 , 21 In response, the World Health Organization (WHO) and the International Agency for Research on Cancer (IARC) have emphasized the need for standardized and comparable evidence that integrates epidemiologic burden with economic valuation to inform cancer control priorities across diverse income settings. To address this gap, a unified welfare-based framework is required to translate health losses into transparent and comparable economic metrics, enabling systematic comparison across diseases, age groups, regions, and income levels. Such an approach can complement conventional epidemiological metrics by identifying settings and disease categories in which welfare losses are relatively concentrated, thereby informing priority-setting for more detailed health-economic evaluation. For example, Pan et al. synthesized GBD 2021 and GLOBOCAN 2022 to describe long-term patterns in incidence, mortality, and DALYs for major hematologic malignancies and projected age-standardized rates through 2040. 22 However, epidemiological metrics are not readily interpretable in macroeconomic terms, and cross-country comparisons of the economic pressure imposed by hematologic malignancies remain limited. In particular, translating disease burden into welfare-based economic valuation is essential to quantify the magnitude of societal loss attributable to both premature mortality and non-fatal health outcomes in a standardized, comparable manner.

Therefore, using standardized data from the Global Burden of Disease Study for 2015–2023, this study applies a Value of Lost Welfare framework to estimate the welfare-based economic burden of major hematologic malignancies across 188 countries. 23 We examine heterogeneity by disease subtype, age, sex, country, and income level, and express monetized health losses relative to national economic output. Supplementary analyses based on WHO Global Health Estimates are used as a descriptive consistency check, while long-horizon extrapolation is treated as exploratory and placed outside the main empirical interpretation. Together, these analyses aim to provide a more transparent comparative account of welfare loss attributable to hematologic malignancies.

Results

Global economic burden of hematologic malignancies

At the global level, leukemia-related VLW increased from US$1.40 trillion in 2015 to US$1.70 trillion in 2023, corresponding to a 21.4% increase, whereas leukemia-related VLW/GDP declined from 1.89% to 1.62%. Among individual disease categories, NHL remained the largest contributor, with VLW increasing from US$1.22 trillion to US$1.46 trillion while VLW/GDP declined from 1.64% to 1.39%. AML was the leading leukemia subtype, with VLW rising from US$0.67 trillion to US$0.80 trillion and VLW/GDP declining from 0.91% to 0.76%. MM also increased in absolute VLW, from US$0.60 trillion to US$0.74 trillion ( Table 1 ). Table 1 Global value of lost welfare (VLW) and VLW/GDP for major hematologic malignancies, 2015 and 2023 Disease category VLW 2015, US$ trillion VLW 2023, US$ trillion VLW change VLW/GDP 2015 VLW/GDP 2023 VLW/GDP change Leukemia 1.4 1.7 21.40% 1.89% 1.62% −14.1% Acute lymphoblastic leukaemia 0.25 0.29 14.00% 0.34% 0.27% −19.3% Acute myeloid leukaemia 0.67 0.8 19.00% 0.91% 0.76% −15.8% Chronic lymphocytic leukaemia 0.21 0.25 17.30% 0.28% 0.23% −17.0% Chronic myeloid leukaemia 0.09 0.12 29.90% 0.12% 0.11% −8.1% Hodgkin lymphoma 0.11 0.12 4.20% 0.15% 0.11% −26.3% Multiple myeloma 0.6 0.74 22.90% 0.81% 0.70% −13.0% Non-Hodgkin lymphoma 1.22 1.46 19.80% 1.64% 1.39% −15.2% Other leukaemia 0.17 0.25 42.10% 0.23% 0.23% 0.60% Values are PPP-adjusted and rounded for readability. VLW/GDP values are expressed as percentages of GDP. Changes represent percentage changes from 2015 to 2023. ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CLL, chronic lymphocytic leukemia; CML, chronic myeloid leukemia; HL, Hodgkin lymphoma; MM, multiple myeloma; NHL, non-Hodgkin lymphoma. Detailed yearly and income-group estimates are provided in Table S1 , and country-level estimates are provided in Table S3 .

Supplementary descriptive comparisons using WHO Global Health Estimates (2000–2021) showed broadly similar directional patterns at the global level ( Figures S1 and S2 ; Table S2 ). These comparisons were used only to assess whether the broad age-, sex-, and disease-specific distributions were directionally consistent across data sources; they were not combined with GBD estimates and were not used to validate the primary model formally.

Trends in the VLW distribution of hematologic malignancies across income levels

Between 2015 and 2023, the composition of VLW attributable to hematologic malignancies exhibited distinct income-specific patterns ( Figure 1 ; Table S1 ). NHL consistently accounted for the largest share of VLW across all income strata. While its proportion remained relatively stable in high-income and upper-middle-income settings, a pronounced upward trend was observed in lower-middle-income and low-income regions. AML also constituted a substantial share of overall VLW. After an initial plateau in high-income countries, its relative contribution declined across other income categories, with the steepest decrease seen in low-income regions. By contrast, CLL, CML and HL each represented small, temporally invariant proportions irrespective of income level. ALL displayed marked income-related heterogeneity: Its VLW share stayed consistently below 5% in high-income regions but was substantially higher in lower-income settings, alongside a declining trajectory across all income strata over time. MM showed the opposite pattern, accounting for over 20% of VLW in high-income regions with minimal temporal fluctuation. Though its contribution remained comparatively low in other income groups, a clear upward trend was evident over the study period. The category of other leukemia maintained a low overall contribution with modest variation across income groups; nevertheless, a moderate increase was noted in low-income and lower-middle-income regions in 2023. Figure 1 Composition of value of lost welfare attributable to hematologic malignancies across income groups, 2015–2023 Data were derived from the global burden of disease (GBD) Study 2015–2023. VLW estimates were calculated by monetizing disability-adjusted life years (DALYs) using the welfare-based VLW framework, adjusted for purchasing power parity (PPP) and income elasticity by World Bank income tier. Each panel A-E illustrate the proportional composition of annual VLW across major hematologic malignancies (Acute lymphoid leukemia , Acute myeloid leukemia , chronic lymphoid leukemia , Chronic myeloid leukemia, Hodgkin lymphoma, Non-Hodgkin lymphoma, Multiple myeloma, and other leukemia) within each income group for the years 2015, 2017, 2019, 2021, and 2023, respectively.

Economic burden of hematologic malignancies by income group and time trends

Distinct temporal patterns in the relative welfare burden of hematologic malignancies were observed across income strata ( Figure 2 ; Table S1 ). Except for HL, other leukemias, and ALL, most hematologic malignancies imposed a higher VLW/GDP burden in high-income settings than in other regions, generally following a declining trajectory over time. Notably, VLW/GDP attributable to AML, CLL, MM, and NHL increased progressively with rising income level. In low-income regions, VLW/GDP contributions from all hematologic malignancies remained comparatively low. In lower-middle-income countries, chronic myeloid leukemia, Hodgkin lymphoma, and “other leukemias” exhibited disproportionately elevated VLW/GDP levels. Among these, chronic myeloid leukemia and “other leukemias” increased appreciably over time, whereas Hodgkin lymphoma followed a marked declining trend. In upper-middle-income regions, VLW/GDP ratios of ALL and other leukemia exceeded those observed in other income categories, with other leukemia displaying a fluctuating upward pattern throughout the study period. Figure 2 VLW/GDP attributable to hematologic malignancies across World Bank income groups, 2015–2023 VLW/GDP represents the ratio of monetized welfare loss (VLW) to gross domestic product (GDP, PPP-adjusted) and is presented as an indicator of relative macroeconomic pressure within the welfare-based valuation framework. Bars show annual VLW/GDP estimates for major hematologic malignancies across World Bank income categories in 2015, 2017, 2019, 2021, and 2023. Panels correspond to disease subtype. Higher values indicate a larger relative welfare burden compared with national economic output.

Economic burden of hematologic malignancies by VLW/GDP across countries

Country-level VLW/GDP ratios showed pronounced heterogeneity across disease categories and geographic settings ( Figure 3 ; Table S3 ). In 2023, the highest leukemia-related VLW/GDP ratios were observed in Bolivia (10.77%), Belarus (4.50%), Thailand (4.27%), Iran (3.83%), and Lebanon (3.77%). For multiple myeloma, the leading ratios were observed in Bolivia (4.30%), Belarus (2.41%), Trinidad and Tobago (1.93%), Latvia (1.83%), and Lithuania (1.67%). Hodgkin lymphoma-related VLW/GDP was highest in Bolivia (1.50%), Georgia (0.79%), Belarus (0.59%), Ukraine (0.54%), and Moldova (0.53%). For non-Hodgkin lymphoma, the highest ratios were observed in Bolivia (8.52%), Brunei Darussalam (3.39%), Belarus (3.34%), Dominica (3.06%), and Slovenia (2.80%). These rankings indicate that high country-level VLW/GDP values were not confined to a single disease category or region, and that some countries, particularly Bolivia and Belarus, appeared repeatedly among the highest-ranked settings. Figure 3 Country-level distribution of VLW/GDP attributable to selected hematologic malignancies in 2023 Country-level VLW/GDP ratios were computed using Global Burden of Disease (GBD) 2023 data and PPP-adjusted GDP. Figure panels A–D separately display the national VLW/GDP percentage ratios for leukemia, multiple myeloma, Hodgkin lymphoma, and non-Hodgkin lymphoma, respectively. Color gradients represent grouped VLW/GDP values presented as percentages of national GDP. Gray shading denotes countries with missing VLW/GDP estimates or mismatched geographic map polygons. The VLW/GDP metric serves as a relative welfare-based disease burden indicator and should not be interpreted as a direct estimate of clinical treatment expenditures or national fiscal impacts.

Temporal changes from 2015 to 2023 were also heterogeneous rather than uniformly increasing or decreasing ( Figure 4 ; Table S3 ). The largest upward shifts in leukemia-related VLW/GDP were observed in Bolivia (+8.41 percentage points), Lebanon (+2.65 percentage points), Tunisia (+1.83 percentage points), Iran (+1.37 percentage points), and Turkiye (+1.05 percentage points), whereas the largest downward shifts occurred in Ukraine (−4.84 percentage points), Bulgaria (−3.59 percentage points), Lithuania (−2.46 percentage points), Moldova (−2.30 percentage points), and Hungary (−2.18 percentage points). For multiple myeloma, the largest increases were seen in Bolivia (+3.84 percentage points), Tunisia (+0.92 percentage points), and Andorra (+0.62 percentage points), while the largest decreases occurred in Ukraine (−1.18 percentage points), Lithuania (−0.95 percentage points), and Czechia (−0.81 percentage points). For non-Hodgkin lymphoma, Bolivia showed the largest increase (+6.68 percentage points), followed by Botswana (+0.66 percentage points), Turkiye (+0.64 percentage points), and Suriname (+0.61 percentage points), whereas Ukraine (−3.00 percentage points), Bulgaria (−2.38 percentage points), and Andorra (−2.17 percentage points) showed the largest decreases. For Hodgkin lymphoma, Bolivia again showed the largest increase (+1.28 percentage points), while Ukraine (−0.82 percentage points) and Moldova (−0.66 percentage points) showed the largest decreases. These country-level results should be interpreted descriptively. A high VLW/GDP ratio may reflect greater modeled health loss, a smaller GDP denominator, demographic structure, or a combination of these factors, rather than direct treatment expenditure, budget impact, or a specific policy effect. Rankings for small economies or countries with limited GDP denominators should therefore be interpreted with particular caution. Figure 4 Country-level change in VLW/GDP attributable to selected hematologic malignancies between 2015 and 2023 Maps show changes in country-level VLW/GDP ratios between 2015 and 2023 for leukemia, multiple myeloma, Hodgkin lymphoma, and non-Hodgkin lymphoma. Values represent percentage-point differences calculated as VLW/GDP in 2023 minus VLW/GDP in 2015. Blue indicates a decrease, orange indicates an increase, and near-neutral shades indicate minimal change. Gray indicates countries without available estimates or unmatched map polygons. VLW/GDP should be interpreted as a relative welfare-based burden indicator rather than a direct measure of treatment expenditure or national budget impact.

Age and sex differences in the economic burden of hematologic malignancies

Age- and sex-specific analyses showed that VLW was generally higher in males than in females across most age strata, while the difference narrowed or reversed in the oldest groups ( Figure 5 ; Table S4 ). Total VLW peaked around ages 60–64 years and then declined in later life. Across disease categories, leukemia contributed a relatively larger share in younger age groups, whereas lymphomas and multiple myeloma accounted for a larger share in older age groups. Supplementary descriptive comparisons based on WHO GHE data showed broadly similar age- and sex-specific distributions ( Figure S3 ). These age patterns should be interpreted as the combined result of disease burden, population age structure, remaining healthy life expectancy, and the age-weighting/discounting embedded in the VSLY function, rather than as a standalone basis for age-prioritized policy decisions. Figure 5 Age- and sex-specific distribution of VLW attributable to hematologic malignancies in 2023 Bars indicate total VLW, expressed in billion US dollars, stratified by sex and 5-year age intervals. Age-specific welfare losses were derived using discounted and age-weighted VSLY functions.

Sensitivity analysis of VLW/GDP

Across all combinations of income elasticity and discount rate, the direction of the main VLW/GDP patterns remained consistent across income strata ( Table S5 ). Increasing the discount rate from 1% to 5% reduced global VLW estimates by approximately 20–30%, with the largest effect in high-income countries. Changes in income elasticity had a smaller effect, generally shifting VLW estimates by approximately 10–15%. These findings indicate that absolute valuation levels depend on parameter assumptions, particularly discounting, whereas the broad comparative patterns were less sensitive to the tested alternatives.

Discussion

This study estimates the welfare burden of hematologic malignancies across countries, income groups, age strata, sex, and disease categories. The central finding is a divergence between absolute and relative burden: Absolute VLW increased from 2015 to 2023, whereas VLW/GDP declined overall. This does not imply that the health impact of hematologic malignancies has diminished. Rather, it indicates that monetized health loss increased in absolute terms, while growth in national economic output lowered the relative ratio in many settings. Because VLW is derived from modeled DALY-based health loss rather than expenditure records, it should be interpreted as a comparative welfare indicator, not as observed healthcare spending. These findings should be viewed in the context of recent GBD-based assessments of hematologic malignancy burden. Li et al. estimated the global, regional, and national economic burden of hematologic malignancies from 1990 to 2021 and projected trends to 2050, reporting a substantial increase in global burden and marked variation across regions and income groups. 7 Other recent work has similarly shown that hematologic malignancies remain a major global burden, with clear differences by subtype, age, sex, and geography. 5 , 8 , 22 The present study extends this literature by focusing on recent GBD 2023 estimates from 2015 to 2023 and by emphasizing VLW/GDP as a relative welfare-burden indicator. This distinction matters because countries with the largest absolute burden are not necessarily those with the greatest burden relative to national economic output. Our results therefore complement prior long-horizon burden assessments by identifying settings where hematologic malignancies impose disproportionate welfare pressure relative to economic capacity.

NHL contributed the largest absolute VLW among the disease categories examined. This is clinically plausible because NHL represents a broad and heterogeneous group of lymphoid malignancies, ranging from aggressive diseases requiring intensive therapy to indolent subtypes managed over long periods. The increase in absolute VLW, despite a decline in VLW/GDP, suggests that NHL remained a major source of welfare loss even as its burden relative to GDP decreased. This welfare signal should not be conflated with changes in lymphoma treatment spending. The therapeutic and cost landscape of B-cell NHL has changed substantially, with the introduction of CD20-directed antibodies, bispecific antibodies, cellular therapies, and rituximab biosimilars. 10 , 12 , 14 , 24 In particular, recent studies have documented uptake of rituximab biosimilars in Medicare and Medicaid and lower prices for oncology biosimilars in China. 24 , 25 These studies are important for health-system budgeting, but they address drug prices, market uptake, and treatment costs. By contrast, the present analysis monetizes modeled disease-related health loss. The NHL results therefore indicate welfare burden from mortality and disability, not lymphoma drug expenditure.

AML accounted for the largest component of leukemia-related VLW and showed only a modest decline in VLW/GDP. This is consistent with AML as an aggressive disease that disproportionately affects older adults, often requires intensive inpatient care or prolonged lower-intensity therapy, and carries substantial relapse-related morbidity and mortality. Claims-based and SEER-Medicare studies have reported high healthcare resource use and costs across induction, remission, and relapse phases, particularly among older adults. 13 , 26 Economic evaluations of venetoclax-based regimens further show that the interpretation of modern AML therapy depends on payer perspective, drug pricing, survival benefit, and willingness-to-pay thresholds. 27 , 28 These clinical and economic studies provide useful context, but they should not be conflated with our estimates. This study does not measure treatment uptake, regimen cost, reimbursement, or line-specific expenditure. Instead, it shows that AML remains a major source of welfare loss after DALY-based health loss is monetized in a standardized cross-country framework. This distinction is especially relevant to low- and middle-income settings, where recent AML therapeutic advances may not translate into comparable access because of diagnostic, infrastructure, and affordability constraints. 29 The category of “other leukemias” showed a relatively large proportional increase in absolute VLW and a broadly stable VLW/GDP ratio. This category is intrinsically heterogeneous and may be influenced by differences in disease classification, registry completeness, and diagnostic resolution across countries. Its results are therefore useful as a broad burden signal, but should not be over-interpreted biologically or clinically without more granular subtype-level data.

MM showed increasing absolute VLW and a declining VLW/GDP ratio. Its larger contribution in high-income countries is consistent with the age distribution of MM and the growing use of long-term, multi-line therapy. Real-world studies have shown that patients with heavily pretreated or triple-class-exposed MM continue to incur substantial healthcare costs, largely related to anti-myeloma drugs, administration, and recurrent healthcare use. 30 , 31 These cost studies do not directly explain VLW, but they support the broader interpretation that MM is becoming a more prominent component of hematologic malignancy burden as populations age and survival improves. CML and HL contributed smaller shares of global VLW, but a small global share does not imply negligible local burden. For CML, imatinib transformed long-term disease control, and generic imatinib has changed the affordability and treatment landscape in some settings. 32 For HL, long-term declines in mortality in many high-income settings are consistent with its smaller global welfare contribution, although regional differences remain. 33 ALL showed a distinct pattern, with a greater relative contribution in lower-income settings. This likely reflects the combination of earlier age at disease onset, differences in treatment access, and the larger welfare consequences of health loss occurring at younger ages under the VSLY framework. Together, these patterns support treating hematologic malignancies as a group of distinct diseases rather than as a single homogeneous burden category.

Age- and sex-stratified analyses showed that VLW was concentrated in mid-to-late working ages and early older age, peaking around ages 60–64 years. This pattern reflects the combined influence of disease incidence, age-specific DALYs, remaining healthy life expectancy, population age structure, and the age-weighted VSLY function. It is also consistent with the epidemiology of hematologic malignancies, in which leukemia and lymphoma subtypes follow distinct age-specific rather than uniform life-course distributions. 8 Men generally had higher VLW than women before advanced age, whereas women exceeded men in the oldest age groups. This pattern is consistent with prior cancer epidemiology studies showing higher incidence, mortality, or poorer survival among men for many cancers, while also reflecting longer female survival into advanced age. 34 The reversal in the oldest age strata should therefore not be interpreted as a simple biological reversal of disease risk. It is more likely to reflect the combined effects of age structure, survival, and valuation assumptions. These age-specific VLW estimates should not be read as a normative allocation rule; they identify where welfare loss is concentrated under the model used here.

Country-level analyses showed that high VLW/GDP ratios were concentrated in selected middle-income and small-economy settings rather than being evenly distributed across regions. This pattern differs from rankings based on absolute burden alone. Previous global economic-burden studies of hematologic malignancies have emphasized high absolute costs in countries with large populations or high aggregate economic output. 7 By contrast, VLW/GDP identifies countries where modeled welfare loss is large relative to national economic capacity. The prominence of Bolivia and Belarus in this analysis therefore does not mean that they contribute the largest share of global VLW; it indicates that their welfare burden is high after scaling VLW to GDP. These country-level findings should be interpreted cautiously. A high VLW/GDP ratio may reflect greater modeled disease burden, a smaller GDP denominator, demographic structure, valuation assumptions, or a combination of these factors. The analysis was not designed to attribute country-level differences to environmental exposure, treatment access, reimbursement systems, procurement policies, or political events. Such explanations would require country-specific registry data, clinical treatment data, expenditure accounts, and policy analyses. The country-level estimates should therefore be viewed as screening signals for more granular investigation, not as causal explanations.

Exploratory supplementary projections suggested that absolute VLW may continue to increase through 2050 across major hematologic malignancy categories, with NHL remaining the largest contributor. Projected VLW/GDP generally declined, echoing the observed 2015–2023 divergence between rising absolute welfare loss and falling relative welfare burden. These projections are useful for showing how historical aggregate patterns may extend under simple time-series assumptions, but they should be interpreted cautiously. They do not incorporate future changes in diagnosis, survival, treatment uptake, drug pricing, biosimilar or generic entry, or financing policy, and should not be read as forecasts of treatment expenditure, budget impact, or health-system strain ( Figures S4 and S5 ).

Taken together, these findings show that welfare-based valuation can add a complementary perspective to hematologic malignancy burden assessment. Epidemiological indicators quantify incidence, mortality, prevalence, and DALYs; direct cost studies measure expenditure within specific health systems; and cost-effectiveness analyses evaluate the value of particular interventions under defined assumptions. VLW/GDP serves a different purpose. It translates modeled health loss into a comparative welfare metric and scales it to national economic output. Used carefully, this metric can help identify disease categories, age strata, and countries where hematologic malignancies generate disproportionate welfare burden and where more detailed clinical-economic evaluation is warranted. The value of this framework is not that it replaces epidemiological surveillance, health-system cost accounting, or formal cost-effectiveness analysis. Rather, it helps connect these lines of evidence. Future studies should combine VLW-based comparative estimates with registry-level clinical data, claims or expenditure records, treatment-access indicators, and policy data. Such integrated analyses will be needed to determine whether high welfare burden reflects disease incidence, survival, treatment availability, out-of-pocket spending, reimbursement design, or broader socioeconomic constraints.

Limitations of the study

Several limitations should be acknowledged. First, this analysis relied on modeled estimates from publicly available datasets and could not capture country-specific reimbursement arrangements, out-of-pocket spending, procurement systems, or real-world treatment intensity. Second, broad disease categories such as AML and NHL were not stratified by molecular subtype, disease phase, treatment line, or regimen, limiting inference about the economic consequences of precision therapies and cellular therapies. Third, World Bank income groups summarize countries with substantial internal heterogeneity and do not capture regional differences in diagnosis, access, or financing. Fourth, the 2015–2023 observation window is relatively short and includes years during which GDP, diagnosis, and health-service use may have been affected by external shocks. Finally, sensitivity analyses varied income elasticity and discounting but did not fully propagate uncertainty from DALY estimation, HALE estimation, or age-specific valuation. The explanatory interpretations in this manuscript should therefore be understood as contextual and descriptive rather than causal. Future work linking burden estimates with claims, expenditure accounts, registry-level clinical data, and treatment-access indicators would be needed to produce more actionable policy estimates.

Resource availability

Lead contact

Further information and requests for resources should be directed to and will be fulfilled by the lead contact, Dengju Li ( lidengju@tjh.tjmu.edu.cn ).

Materials availability

This study did not generate new unique reagents or biological materials.

Data and code availability

• Data: The input data used in this study are publicly available from the Global Burden of Disease Results Tool, GBD life-table outputs, the WHO Global Health Estimates database, and the World Bank World Development Indicators. Processed summary tables supporting the main and supplementary analyses are provided in Tables 1 and S1 , S2 , S3 , S4 , S5 , and S6 . • Code: Analytic scripts, figure-generation code, and session information are provided as Data S1 and have been deposited at Zenodo: https://doi.org/10.5281/zenodo.20501586 . • Any additional information: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This study was financially supported by the 10.13039/501100001809 National Natural Science Foundation of China (grant no. 82270177 ) and the Joint Fund for Innovation and Development of the 10.13039/501100003819 Hubei Provincial Natural Science Foundation (grant nos. 2024AFD422 and 2025AFD788 ). We thank the Global Burden of Disease (GBD) Study for providing the data used in this research. We also acknowledge the World Bank World Development Indicators for the GDP data used in our analysis.

Author contributions

D.J.L. contributed to the conceptualization and design of the study. Y.M.Z. was responsible for data collection and verification. X.Y.C. prepared the manuscript and created the figures. Y.M.Z. and Y.Y.Z. carried out data analysis and interpretation. All authors contributed to the manuscript revision. Y.M.Z. and X.Y.C. made equal contributions to this work. The final manuscript was reviewed and approved by all authors.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE SOURCE IDENTIFIER GBD 2023 Results Tool Institute for Health Metrics and Evaluation https://vizhub.healthdata.org/gbd-results/ GBD life-table outputs Institute for Health Metrics and Evaluation https://ghdx.healthdata.org/ WHO Global Health Estimates World Health Organization https://www.who.int/data/global-health-estimates World Bank World Development Indicators World Bank https://databank.worldbank.org/source/world-development-indicators Analytic scripts and session information This paper Data S1 ; Zenodo: https://doi.org/10.5281/zenodo.20501586 R version 4.3.2 R Project https://www.r-project.org/ dplyr CRAN https://cran.r-project.org/package=dplyr tibble CRAN https://cran.r-project.org/package=tibble readr CRAN https://cran.r-project.org/package=readr forecast CRAN https://cran.r-project.org/package=forecast ggplot2 CRAN https://cran.r-project.org/package=ggplot2 maps CRAN https://cran.r-project.org/package=maps

Experimental model and study participant details

Omitted as this study does not involve biological models, animal experiments, cell lines, primary cell cultures, human participant recruitment, patient-level samples, or clinical trial participants. Analyses were based entirely on publicly available, deidentified, aggregated data. Sex-stratified estimates were analyzed where available and are reported in the Results, Figure 5 , and Table S4 . Gender, ancestry, race, and ethnicity were not available in the aggregate datasets used in this study.

Method details

Data source

Data for the primary analyses were obtained from the Global Burden of Disease (GBD) 2023 Results Tool ( https://vizhub.healthdata.org/gbd-results/ ). We extracted estimates for 2015, 2017, 2019, 2021, and 2023 and focused on eight hematologic malignancy categories: acute lymphoblastic leukemia, acute myeloid leukemia, chronic lymphocytic leukemia, chronic myeloid leukemia, Hodgkin lymphoma, non-Hodgkin lymphoma, multiple myeloma, and other leukemia. 35 DALYs were extracted as absolute counts and processed at the country-age-sex-cause-year stratum before any aggregation. Remaining healthy life expectancy (HALE), defined as expected years lived in full health at the midpoint of each 5-year GBD age group, was obtained from GBD/IHME life-table outputs. Annual country-level gross domestic product (GDP) and GDP per capita, both adjusted for purchasing power parity (PPP), were obtained from the World Bank World Development Indicators and matched by calendar year. The main analysis was restricted to 2015-2023 to emphasize recent empirically observed burden patterns rather than long-range projections.

Analytical workflow

The analysis proceeded in five steps. First, DALYs were harmonized by country, age group, sex, disease category, and calendar year. Second, each country-year stratum was linked to PPP-adjusted GDP and GDP per capita. Third, country-specific values of a statistical life and age-specific annual values of life-years were estimated using income transfer, age-weighting, and discounting assumptions. Fourth, DALYs were monetized at the most granular available stratum before aggregation. Fifth, VLW estimates were summarized globally and by country, World Bank income group, disease subtype, age, and sex; VLW/GDP was then calculated as a relative indicator of welfare-based macroeconomic pressure. Sensitivity analyses were used to evaluate whether the main patterns were sensitive to income-elasticity and discount-rate assumptions.

Calculation of the value of lost welfare

The economic valuation of health losses was conducted using a welfare-based Value of Lost Welfare (VLW) framework adapted from prior global health economic applications. This approach was selected because DALYs capture fatal and non-fatal health loss but do not represent payments, claims, or treatment expenditure. A direct cost-of-illness framework would therefore require country-level expenditure, reimbursement, and treatment-intensity data that are not consistently available across settings. VLW instead monetizes health loss in a standardized welfare framework and is intended for comparative burden assessment. For each country c and calendar year y, the country-specific peak VSL was estimated as: VSLpeak , c , y = VSLpeak , USA , 2019 × ( GDPpc , c , y / GDPpc , USA , 2019 ) ∧ IEc

Where GDPpc,c,y denotes the year-specific per-capita PPP GDP of country c, and IEc denotes the income elasticity used to transfer the reference VSL across income settings. 36 We anchored VSLpeak,USA,2019 at US$9,979,014 (2019 USD), based on USDA Economic Research Service estimates. The anchor was held constant to provide a transparent reference point, while GDP per capita varied by country and year. In the primary analysis, income elasticities were assigned by World Bank income tier (high income = 1.0; upper-middle income = 1.2; lower-middle income = 1.3; low income = 1.5), reflecting the common practice of allowing valuation transfer to vary more strongly in lower-income settings.

Lifetime VSLs were converted to age-specific annual values using an age-weighting function, f(a) = max(1 – 0.001 × (a – 40) 2 , 0.2), and discounting of remaining healthy life years at 3% per annum. The age-weighting function was used to distribute lifetime value across age groups rather than to assign normative priority to any age group. The 3% discount rate was used as the base-case health economic convention, and alternative rates were evaluated in sensitivity analyses: VSLYc , a , y = VSLpeak , c , y × f ( a ) / Σ t > = a [ f ( t ) × 5 × ( 1 + r ) ∧ ( t − a ) ]

Valuation preceded aggregation. For each country, age, sex, cause stratum: VLWc , a , s , d , y = DALYc , a , s , d , y × VSLYc , a , y

Where c denotes country, a age group, s sex, d cancer subtype, and y calendar year. Valuation was performed before aggregation so that age, sex, country, and disease-specific differences in DALYs and valuation assumptions were preserved. Stratum-level VLW estimates were then aggregated to country, income-group, and global levels. VLW/GDP ratios were computed as total VLW divided by year-specific PPP-adjusted GDP and should be interpreted as relative welfare-based economic pressure rather than literal macroeconomic loss, direct healthcare spending, or national budget impact. 37 , 38 A consolidated summary of equations, assumptions, variable definitions, and aggregation steps is provided in Table S6 .

Exploratory forecasting analysis

Long-horizon projections were retained only as exploratory supplementary analyses. For this component only, annual disease-specific VLW series for 1990-2023 were reconstructed using the same DALY-to-VLW valuation procedure and then extended to 2050 using log-transformed time-series extrapolation with automated stepwise ARIMA modeling; exponential smoothing was used when a constant model was selected. Because reliable country-specific projections of DALYs by hematologic subtype were not available, these analyses were designed as descriptive extensions of historical series rather than causal forecasts of treatment expenditure, drug uptake, or future health-system strain. The assumptions and outputs of this component are reported in the Supplementary Materials, including exploratory projections of absolute VLW and VLW/GDP in Figures S4 and S5 , and are not used as primary evidence in the main interpretation.

Supplementary descriptive consistency check using Global Health Estimates

In addition to the GBD-based primary analyses, supplementary descriptive analyses were conducted using data from the World Health Organization Global Health Estimates (GHE) database to assess the extent to which broad directional patterns were consistent across sources. GHE data were not combined with GBD estimates in model construction or forecasting. Instead, they were used solely as descriptive consistency checks for age-, sex-, and disease-specific VLW distributions at the global level for selected years between 2000 and 2021.

Quantification and statistical analysis

All analyses were conducted in R (version 4.3.2) using dplyr, tibble, readr, forecast, and ggplot2. Fixed random seeds were used where applicable to improve reproducibility. Ethics approval was not required because the study relied entirely on publicly available, de-identified, aggregated data.

Sensitivity analysis

Sensitivity analyses jointly varied income elasticity and the discount rate to evaluate the dependence of VLW estimates on key valuation assumptions. Two alternative IE sets were assessed relative to the base case: a low-elasticity set (high income = 0.8; upper-middle income = 1.0; lower-middle income = 1.2; low income = 1.3) and a high-elasticity set (high income = 1.2; upper-middle income = 1.4; lower-middle income = 1.5; low income = 1.7). Discount rates of 1% and 5% were evaluated cross-factorially against these IE sets, generating a 3 × 3 grid (IE_low/IE_base/IE_high × r = 0.01/0.03/0.05), while holding the US anchor VSL and HALE surfaces fixed.

References

1 The Lancet Healthy, L. Ageing populations: unaffordable demography Lancet Healthy Longev. 3 2022 e804 10.1016/s2666-7568(22)00272-0 36480979

2 Hong C. Sun L. Liu G. Guan B. Li C. Luo Y. Response of Global Health Towards the Challenges Presented by Population Aging China CDC Wkly. 5 2023 884 887 10.46234/ccdcw2023.168 37814614 PMC10560387

3 Lee-Six H. Øbro N.F. Shepherd M.S. Grossmann S. Dawson K. Belmonte M. Osborne R.J. Huntly B.J.P. Martincorena I. Anderson E. Population dynamics of normal human blood inferred from somatic mutations Nature 561 2018 473 478 10.1038/s41586-018-0497-0 30185910 PMC6163040

4 Mitchell E. Spencer Chapman M. Williams N. Dawson K.J. Mende N. Calderbank E.F. Jung H. Mitchell T. Coorens T.H.H. Spencer D.H. Clonal dynamics of haematopoiesis across the human lifespan Nature 606 2022 343 350 10.1038/s41586-022-04786-y 35650442 PMC9177428

5 Zhang N. Wu J. Wang Q. Liang Y. Li X. Chen G. Ma L. Liu X. Zhou F. Global burden of hematologic malignancies and evolution patterns over the past 30 years Blood Cancer J. 13 2023 82 10.1038/s41408-023-00853-3 37193689 PMC10188596

6 Skirbekk V. Dieleman J.L. Stonawski M. Fejkiel K. Tyrovolas S. Chang A.Y. The health-adjusted dependency ratio as a new global measure of the burden of ageing: a population-based study Lancet Healthy Longev. 3 2022 e332 e338 10.1016/s2666-7568(22)00075-7 36098308

7 Li Z. Chen G. Du G. Global, regional, and national economic burden of hematologic malignancies (1990-2021) with projections to 2050 Front. Public Health 13 2025 1570792 10.3389/fpubh.2025.1570792 PMC12245848 40655204

8 Daltveit D.S. Morgan E. Colombet M. Steliarova-Foucher E. Bendahhou K. Marcos-Gragera R. Rongshou Z. Smith A. Wei H. Soerjomataram I. Global patterns of leukemia by subtype, age, and sex in 185 countries in 2022 Leukemia 39 2025 412 419 10.1038/s41375-024-02452-y 39567675

9 Berhan A. Almaw A. Damtie S. Solomon Y. Diffuse large B cell lymphoma (DLBCL): epidemiology, pathophysiology, risk stratification, advancement in diagnostic approaches and prospects: narrative review Discov. Oncol. 16 2025 184 10.1007/s12672-025-01958-w 39954204 PMC11829893

10 Dickinson M.J. Carlo-Stella C. Morschhauser F. Bachy E. Corradini P. Iacoboni G. Khan C. Wróbel T. Offner F. Trněný M. Glofitamab for Relapsed or Refractory Diffuse Large B-Cell Lymphoma N. Engl. J. Med. 387 2022 2220 2231 10.1056/NEJMoa2206913 36507690

11 Bröckelmann P.J. Cliff E.R.S. Iacoboni G. Simon F. Horowitz M.M. Keating A. Mateos M.V. Mohty M. Sidana S. Song Y. Beyond maximum grade: tolerability of immunotherapies, cellular therapies, and targeted agents in haematological malignancies Lancet Haematol. 12 2025 e470 e481 10.1016/s2352-3026(25)00051-1 40447355

12 Anderson G.S.F. Chapman M.A. T cell-redirecting therapies in hematological malignancies: Current developments and novel strategies for improved targeting Mol. Ther. 32 2024 2856 2891 10.1016/j.ymthe.2024.07.028 39095991 PMC11403239

13 Choi M. Song J. Bui C.N. Ma E. Chai X. Yin L. Betts K.A. Kapustyan T. Montez M. LeBlanc T.W. Costs per patient achieving remission with venetoclax-based combinations in newly diagnosed patients with acute myeloid leukemia ineligible for intensive induction chemotherapy J. Manag. Care Spec. Pharm. 28 2022 980 988 10.18553/jmcp.2022.22021 35708343 PMC12101567

14 Hernandez I. Prasad V. Gellad W.F. Total Costs of Chimeric Antigen Receptor T-Cell Immunotherapy JAMA Oncol. 4 2018 994 996 10.1001/jamaoncol.2018.0977 29710129 PMC6145722

15 Pettifor J.M. Saloojee H. Unequal access: the prohibitive costs of new drugs in low- and middle-income countries Nat. Rev. Nephrol. 21 2025 281 282 10.1038/s41581-025-00948-z 40087511

16 Sharma T. Kruja K. Årdal C. Lopert R. Mestre-Ferrandiz J. Boselli S. Vogler S. Managing costly new drugs to support equitable access: international collaboration and patient engagement to assess value, and managed entry agreements BMJ (Clinical research ed.) 391 2025 e086174 10.1136/bmj-2025-086174 PMC12599528 41218988

17 Callenbach M.H.E. Vreman R.A. Leopold C. Mantel-Teeuwisse A.K. Goettsch W.G. Managed Entry Agreements for High-Cost, One-Off Potentially Curative Therapies: A Framework and Calculation Tool to Determine Their Suitability Pharmacoeconomics 43 2025 53 66 10.1007/s40273-024-01433-4 39368017 PMC11724790

18 Mahumud R.A. Optimising Cancer Medicine in Clinical Practices: Are Neoadjuvant and Adjuvant Immunotherapies Affordable for Cancer Patients in Low- and Middle-Income Countries? Cancers 17 2025 1722 10.3390/cancers17101722 PMC12109690 40427219

19 Oldfield L. Penm J. Mirzaei A. Moles R. Prices, availability, and affordability of adult medicines in 54 low-income and middle-income countries: evidence based on a secondary analysis Lancet. Glob. Health 13 2025 e50 e58 10.1016/s2214-109x(24)00442-x 39706661

20 Lensberg B.R. Drummond M.F. Danchenko N. Despiégel N. François C. Challenges in measuring and valuing productivity costs, and their relevance in mood disorders Clinicoecon. Outcomes Res. 5 2013 565 573 10.2147/ceor.S44866 24273412 PMC3836685

21 Pike J. Grosse S.D. Friction Cost Estimates of Productivity Costs in Cost-of-Illness Studies in Comparison with Human Capital Estimates: A Review Appl. Health Econ. Health Policy 16 2018 765 778 10.1007/s40258-018-0416-4 30094591 PMC6467569

22 Pan T. Zhang J. Wang X. Song Y. Global burden and trends of hematologic malignancies based on Global Cancer Observatory 2022 and Global Burden of Disease 2021 Exp. Hematol. Oncol. 14 2025 98 10.1186/s40164-025-00684-x 40676693 PMC12273037

23 Alkire B.C. Shrime M.G. Dare A.J. Vincent J.R. Meara J.G. Global economic consequences of selected surgical diseases: a modelling study Lancet. Glob. Health 3 2015 S21 S27 10.1016/s2214-109x(15)70088-4 25926317 PMC4884437

24 Qian J. Uptake of rituximab biosimilars in Medicare and Medicaid in 2019-2022 Am. J. Manag. Care 30 2024 e359 e363 10.37765/ajmc.2024.89644 39745502

25 Luo X. Du X. Li Z. Liu J. Lv X. Li H. Guo Q. Wang C. Xue X. Le K. Clinical Benefit, Price, and Uptake for Cancer Biosimilars vs Reference Drugs in China: A Systematic Review and Meta-Analysis JAMA Netw. Open 6 2023 e2337348 10.1001/jamanetworkopen.2023.37348 PMC10570888 37824143

26 Brunner A.M. Huggar D. Copher R. Zhou Z.Y. Zichlin M.L. Anderson A. Downes N. McBride A. Economic burden during remission and after relapse among older patients with newly diagnosed acute myeloid leukemia without hematopoietic stem cell transplant: A retrospective study using the SEER-Medicare database Leuk. Res. 132 2023 107353 10.1016/j.leukres.2023.107353 37562330

27 Pratz K.W. Chai X. Xie J. Yin L. Nie X. Montez M. Iantuono E. Downs L. Ma E. Cost-Effectiveness Analysis of Venetoclax in Combination with Azacitidine Versus Azacitidine Monotherapy in Patients with Acute Myeloid Leukemia Who are Ineligible for Intensive Chemotherapy: From a US Third Party Payer Perspective Pharmacoeconomics 40 2022 777 790 10.1007/s40273-022-01145-7 35696071 PMC9300490

28 Patel K.K. Zeidan A.M. Shallis R.M. Prebet T. Podoltsev N. Huntington S.F. Cost-effectiveness of azacitidine and venetoclax in unfit patients with previously untreated acute myeloid leukemia Blood Adv. 5 2021 994 1002 10.1182/bloodadvances.2020003902 33591323 PMC7903235

29 Morcos-Sandino M. Quezada-Ramírez S.I. Gómez-De León A. Advances in the Treatment of Acute Myeloid Leukemia: Implications for Low- and Middle-Income Countries Biomedicines 13 2025 1221 10.3390/biomedicines13051221 PMC12109363 40427048

30 Jagannath S. Joseph N. He J. Crivera C. Fu A.Z. Garrett A. Shah N. Healthcare Costs of Multiple Myeloma Patients with Four or More Prior Lines of Therapy, Including Triple-Class Exposure in the United States Oncol. Ther. 10 2022 411 420 10.1007/s40487-022-00198-0 35579821 PMC9681939

31 Yang J. Boytsov N. Carlson J.J. Barthold D. Health care resource utilization and costs among patients with multiple myeloma with exposure to double-class or triple-class multiple myeloma treatments: A retrospective US claims database analysis J. Manag. Care Spec. Pharm. 29 2023 917 926 10.18553/jmcp.2023.29.8.917 37523320 PMC10397331

32 Gorkin L. Kantarjian H. Targeted therapy: Generic imatinib - impact on frontline and salvage therapy for CML Nat. Rev. Clin. Oncol. 13 2016 270 272 10.1038/nrclinonc.2016.59 27098218 PMC5450934

33 Levi F. Lucchini F. Negri E. Boyle P. La Vecchia C. Trends in mortality from Hodgkin's disease in western and eastern Europe Br. J. Cancer 87 2002 291 293 10.1038/sj.bjc.6600452 12177797 PMC2364217

34 Cook M.B. McGlynn K.A. Devesa S.S. Freedman N.D. Anderson W.F. Sex disparities in cancer mortality and survival Cancer Epidemiol. Biomarkers Prev. 20 2011 1629 1637 10.1158/1055-9965.Epi-11-0246 21750167 PMC3153584

35 Global Burden of Disease Collaborative Network Global Burden of Disease Study 2023 (GBD 2023) Results 2025 Institute for Health Metrics and Evaluation (IHME) Seattle, United States https://vizhub.healthdata.org/gbd-results/

36 Robinson L.A. Hammitt J.K. Research Synthesis and the Value per Statistical Life Risk Anal. 35 2015 1086 1100 10.1111/risa.12366 25851065

37 Johansson P.O. The value of a statistical life: theoretical and empirical evidence Appl. Health Econ. Health Policy 1 2002 33 41 14618746

38 Robinson L.A. Hammitt J.K. O'Keeffe L. Valuing Mortality Risk Reductions in Global Benefit-Cost Analysis J. Benefit-Cost Anal. 10 2019 15 50 10.1017/bca.2018.26 32968616 PMC7473065

Supplemental information

Document S1. Figures S1–S5 and Tables S1, S2, S5, and S6 Table S3. Country-level VLW and VLW/GDP estimates for hematologic malignancies, 2015–2023 Table S4. Age- and sex-specific VLW by disease subtype in 2023 Data S1. Analytic scripts, figure-generation code, and session information

Supplemental information can be found online at https://doi.org/10.1016/j.isci.2026.116746 .