Leukemia epidemiology in China: Burden, trends, and determinants in the 21st century

Abstract

This comprehensive review integrates population-based registries, hospital databases and Global Burden of Disease data to describe the evolving leukemia burden in China from 2000 to 2022. The overall incidence has stabilized nationally, but the absolute number of cases continues to increase as the population ages. A bimodal age pattern persists, with acute leukemias clustering in young children and older adults, while chronic forms predominate in mid-to-late life, and males are consistently more affected by all subtypes. Rapid expansion of haploidentical hematopoietic stem cell transplantation has resulted in marked survival gains for both acute myeloid leukemia and acute lymphoblastic leukemia, and its seamless integration with molecularly targeted agents, venetoclax-based regimens and chimeric antigen receptor T-cell therapy has transformed acute leukemias into potentially curable diseases for an expanding proportion of patients. In parallel, universal access to tyrosine kinase inhibitors and standardized molecular monitoring have turned chronic myeloid leukemia into a manageable chronic condition, and survival of patients with chronic lymphocytic leukemia is improving as novel Bruton’s tyrosine kinase and BCL-2 inhibitors diffuse into clinical practice. Tobacco, obesity, benzene and radon remain the principal modifiable drivers of leukemogenesis. Strengthening data completeness, widening equitable access to precision therapies and controlling these environmental risks are essential to sustaining the observed continuous improvement in leukemia patient survival and ensuring that ever more Chinese patients achieve a cure or durable disease control in the decades ahead.

Introduction

Leukemia is a biologically heterogeneous family of hematopoietic stem cell malignancies that are first subdivided by natural history into acute (≥20% blasts, rapid course) and chronic (mature-appearing cells, indolent course) forms and then by lineage into myeloid or lymphoid subtypes. The World Health Organization (WHO) 2022 taxonomy integrates morphology, immunophenotype, cytogenetics and molecular genetics to recognize more than 70 entities (1-4).

Globally, leukemia accounts for approximately 3% of all incident cancers, but it ranks among the five leading causes of cancer death in children and adolescents (5). China, the world’s most populous country, now contributes one-quarter of global leukemia deaths, although it represents only one-sixth of the world population. Leukemia constitutes about 3.5% of all cancer deaths in China (ranking 6th) but is the second most common childhood malignancy and the leading cause of cancer death among adolescents.

Over the past two decades, the country has experienced an unprecedented demographic transition: the proportion of citizens aged ≥60 years rose from 10% in 2000 to 19% in 2022 and is projected to reach 28% by 2035. Parallel rapid industrialization has generated widespread environmental exposures—ambient benzene, formaldehyde, fine particulate matter (PM2.5) and ionizing radiation—which are recognized as leukemogens [International Agency for Research on Cancer (IARC), 2012]. Moreover, China has implemented phased reforms of its health insurance system, achieving near-universal coverage by 2011 and expanding the benefit package to include expensive targeted agents such as imatinib and dasatinib since 2017. These macrolevel changes jointly shape both the epidemiological profile and the observable survival trends of leukemia in the 21st century (6).

The first two decades of this century have witnessed milestone advances in the diagnosis and treatment of hematological malignancies in China. According to the 2019 National Cancer Registry (NCR) data (which cover 628 million people and 919 population-based cancer registries), there were 43,275 new cases of acute leukemia and 27,049 deaths annually, with age-standardized incidence and mortality rates of 2.83 and 1.51 per 100,000 population, respectively, and the incidence showed a bimodal distribution pattern. In this context, therapeutic breakthroughs for various hematological malignancies have been particularly remarkable.

In the treatment of acute leukemia, the 5-year overall survival (OS) rate for pediatric acute lymphoblastic leukemia (ALL) and acute promyelocytic leukemia (APL) patients has reached 85%−91%, while the survival rate remains low for adult acute myeloid leukemia (AML) patients (24%) and patients aged ≥60 years (<15%) (7). The scale of allogeneic hematopoietic stem cell transplantation (allo-HSCT) has expanded significantly, from fewer than 500 procedures in 2000 to more than 21,000 in 2022, making China one of the world’s largest transplant hubs (China HSCT Registry data, 2023). Notably, the application of haploidentical hematopoietic stem cell transplantation (haplo-HSCT) has surpassed that of matched-sibling donor transplantation, accounting for 67% of 9,056 allo-HSCT cases (8-15). With respect to pediatric and adult B-cell ALL (B-ALL), CD19-targeted chimeric antigen receptor (CAR) T-cell therapy has achieved a complete remission rate exceeding 80% in relapsed/refractory patients (16,17).

The treatment of chronic myeloid leukemia (CML) and chronic lymphocytic leukemia (CLL) relies on lineage-specific oncogene-targeted interventions: CML has been transformed into a chronically manageable disease through imatinib (launched in 2002) and subsequent tyrosine kinase inhibitors (TKIs), with a 10-year OS rate exceeding 85%. Among patients with deep remission, 45%−50% can discontinue treatment and maintain treatment-free remission (TFR) under standardized BCR-ABL1 quantitative polymerase chain reaction (qPCR) monitoring (with a threshold of 0.1%) (18-23). For CLL, fixed-duration ibrutinib-venetoclax (VEN) combination therapy has increased the 3-year disease-free survival rate to 74%, challenging the traditional lifelong kinase inhibition model.

However, treatment accessibility remains uneven. Only 40% of rural CML patients receive regular molecular monitoring, and reimbursement for targeted agents in CLL is limited to the 17p-deletion subtype (24). In addition, urban, eastern and high socioeconomic status (SES) populations benefit earlier and more substantially than their rural, western or low-SES counterparts do.

Despite this remarkable clinical progress, nationwide epidemiological intelligence remains fragmented. The NCR of China expanded from 5 local registries in 2000 to 722 in 2022; however, the population coverage is still only 55%, and pathology-based subtyping is available for less than 40% of reported cases (25-27). Hospital-based medical records (HBMRs) contain rich molecular information but lack denominators, making external validity uncertain. Consequently, national estimates of incidence, mortality and disability-adjusted life-years (DALYs) rely heavily on modeled extrapolations rather than empirical enumeration. The quantification of avoidable risk factors is even scarcer. Existing studies are mostly small-scale case-control investigations focused on single exposures such as benzene or smoking, and national-level population-attributable fractions (PAFs) have not been systematically estimated since the Global Burden of Disease (GBD) 2019 iteration.

To address these evidence gaps, this review synthesizes all accessible data sources—NCR, GBD 2021, HBMR, vital statistics and published cohort studies—to characterize the leukemia burden, trends and determinants across mainland China’s 31 provinces between 2000 and 2022.

By integrating population-based registry statistics with high-resolution exposure data and advanced modeling techniques, this review provides the most comprehensive and up-to-date assessment of leukemia epidemiology in China and informs priority setting for cancer control policies in the era of precision hematology.

Incidence

Over the past three decades, the incidence of each major leukemia subtype in China has followed a distinct trajectory shaped by demographic transition, environmental exposure and differential access to targeted therapy.

In China, AML is the most common type of acute leukemia at every age. Its age-standardized incidence rate (ASIR, Segi World Standard) was 1.24 per 100,000 in 2019 in the NCR, which covers 628 million persons; the incidence accelerates after 60 years and peaks at 6.26 per 100,000 in the 75−79-year group. Males outnumber females by about 40% starting in infancy, and the male-to-female ratio widens further beyond the sixth decade (7). Long-term data from GBD 2021 show that the absolute number of incident cases rose by 28% from 1990 to 2021 in parallel with population growth and aging, yet the national ASIR has plateaued at 1.2−1.3 per 100,000 since 2005, indicating that demographic restructuring—rather than new environmental or lifestyle risks—is the dominant driver (28-30). The widening sex differential after midlife mirrors male’s greater lifetime exposure to occupational benzene and tobacco; >70% of AML DALYs in males ≥50 years are attributed to smoking, high body mass index (BMI) or occupational solvent exposure. After adolescence, AML dominates the leukemia landscape: the second hump of the bimodal incidence curve beyond 80 years is generated largely by AML because B-ALL almost disappears after age 25 years (31). Consequently, AML constitutes 74% of leukemia diagnoses in adults >60 years and drives the sex ratio from 1.25 among patients aged 15−49 years old to 1.68 among patients aged over 60 years (31).

Provincial heterogeneity is most pronounced for AML: steel- and petrochemical-heavy provinces such as Guizhou, Gansu and Hebei have ASIRs that are 4−5 times greater than the national average, a gradient that tracks cumulative ambient benzene and PM2.5 better than tobacco use alone does. Urban residency results in a 25% increase in transplant utilization and a 12% decrease in 12-month mortality, whereas rural counties suffer a 1.4-fold increase in early mortality because intensive chemotherapy and transplantation are centralized in tertiary centers. Bayesian age-period-cohort modeling predicts that, without sex-specific tobacco and occupational controls, the male AML excess will increase the sex ratio to 1.77 by 2036 (28-31).

In China, ALL is the fastest-growing hematological malignancy. Age-period-cohort dissection of the GBD 2019 revealed net annual increases of 7.1% in males and 7.5% in females, with statistically significant positive local drifts in every 5-year age group from birth to late adulthood (32). Projections indicate that the ASIR will surge by 64% in males and 75% in females between 2019 and 2028, reaching 6.4 and 6.3 per 100,000, respectively—well above the current global average (32). This steep trajectory is driven by three modifiable exposures: tobacco smoke (polycyclic aromatic hydrocarbons and benzene directly damage hematopoietic stem cell DNA, accounting for >50% of the PAF), rising adult overweight/obesity (which now contributes 30% of leukemia-related DALYs through systemic inflammation and insulin-mediated mitogenic signaling), and residual occupational benzene exposure in small-scale manufacturing and petro-chemical sectors, especially among males born after 1980 (33). Demographic transitions magnify the burden: later-born generations carry a steadily increasing lifetime risk, whereas rapid population aging expands the reservoir of older adults in whom clonal hematopoiesis and cumulative carcinogen exposure converge. Although adults >60 years currently constitute <20% of incident ALL cases, their proportional contribution is expected to rise steeply as the national age pyramid inverts.

Notably, China’s age-sex-geographic heterogeneity in ALL is largely confined to the first two decades of life: the entire early childhood incidence peak (<5 years, about 17.7 per 100,000) is attributable to ALL, whereas the second adolescent/young adult peak is almost exclusively driven by AML (34). Within this age-restricted window, the male-to-female ratio is only 1.1 before puberty but increases sharply to 1.4 thereafter, reflecting the persistence of pediatric-type biology. Provincial variation in childhood ALL incidence is modest (range: 3.8−5.2 per 100,000), but excess adult ALL in petro-chemical provinces is amplified by the same benzene/smoking matrix that underpins AML risk (35). Unless tobacco and obesity trajectories flatten, the post-40 sex gap will widen further. Taken together, these converging forces indicate that the ALL incidence in China will continue to steeply increase, necessitating urgent, targeted prevention strategies focused on tobacco control, obesity reduction, and workplace chemical safety, in addition to age-specific and region-specific care pathways (32,33).

In China, the epidemiologic profile of CML follows a coherent narrative of “biological incidence flat, single-peaked age curve, modest regional scatter, and outcome revolution through policy-plus-technology”. Since 2000, registries covering >40% of the mainland have recorded an ASIR of 0.7−0.9 per 100,000 (provincial range: 0.6−1.1), implying that the true biological occurrence has been neither inflated by better diagnostics nor suppressed by nationwide environmental changes (36). Unlike acute leukemias, CML shows no childhood or late-life surge: the incidence increases monotonically after age 20 years and plateaus only after 70 years, while the male-to-female ratio remains at 1.5−1.6 across all 31 provinces, mirroring the predominance of males in terms of smoking and occupational benzene exposure, which also drives AML (36). Among birth cohorts after 1980, this ratio narrowed from the historical 1.6−1.8 to about 1.3, paralleling a 60% decrease in male smoking prevalence and tighter occupational benzene limits introduced between 2002 and 2015 (37). Although CML is less influenced by the aging-related clonal hematopoiesis of indeterminate potential (CHIP) than acute leukemias are, the same petro-chemical provinces that form the high-AML cluster still rank the highest for CML, albeit with narrower geographic spread (36). The 2017 inclusion of generic imatinib in the National Reimbursement Drug List, coupled with centralized qPCR monitoring, has converted this stable incidence into a public health success story: urban registries report only a 10% higher incidence but a 20% superior major molecular response rate at 12 months, and contemporary real-world cohorts achieve 85% molecular response rates (BCR-ABL1≤0.1% at 12 months) and >85% 10-year OS—figures that match high-income registries and continue to improve with second- and third-generation TKIs (18,20,38). Forward projections to 2036 anticipate that the ASIR will remain stable, but the male excess will persist unless the enforcement of occupational benzene limits is accelerated (33,38,39).

In China, the epidemiologic logic of CLL is best described as “low-level flat, elderly male concentration, access disparity amplified”. The NCR records an ASIR of only 0.3−0.4 per 100,000—approximately one-tenth the Caucasian figure—with no appreciable temporal rise, rendering the classic bimodal age-incidence curve virtually invisible (40). Disease is concentrated in the elderly (median diagnosis age 70 years) and displays the steepest sex gradient of all leukemias (male:female=2:1), mirroring the lifetime cumulative exposure to tobacco and benzene (40). Provincial variation is compressed (0.2−0.5 per 100,000), yet provinces with the highest sociodemographic index and smoking prevalence consistently occupy the top quartile. Reimbursement restricts ibrutinib and VEN to 17p-deleted cases, so bendamustine-based chemoimmunotherapy remains front-line outside tier-1 cities; consequently, urban patients gain earlier access to novel agents and clinical trials, whereas rural patients experience a 1.3-fold higher 24-month treatment failure rate (40,41). Bayesian modeling indicates that unless access to novel agents is broadened, the current modest interprovincial incidence gap will persist, but the pronounced male predominance will remain stable through 2036 (40,41).

Collectively, these subtype-specific trends underscore a leukemia transition in China: AML incidence driven by aging, ALL by aging plus emerging lifestyle/industrial exposures, CML controlled by universal TKI access, and CLL still constrained by a low biological prevalence and uneven diffusion of novel agents.

Treatment advancement and survival

The relative survival rate for patients with leukemia in China has significantly improved over the past decades.

Over the past four decades, AML in China has evolved from a near-uniformly fatal disease—a median survival of about 3 months without therapy and a 5-year relative survival of <20% only two decades ago—to a clinically manageable malignancy. A 2025 nationwide epidemiology study captured the latest acceleration: population-based registries revealed that the 5-year OS of adult AML (non-APL) increased from 23.9% in 2016−2018 to 24.7% in 2019−2020, with gains concentrated in patients <60 years (7). This continuing upward trajectory rests on three parallel breakthroughs: 1) the national transplantation rate increased from 16.7% to 21.8% (7,42). Haplo-HSCT, now accounting for 67% of all transplants in 2023, has become a game changer. The “Beijing Protocol” not only resolved the long-standing donor-shortage bottleneck but also outperformed conventional matched-donor transplantation in patients with high-risk leukemia (43), achieving a 3-year leukemia-free survival of 73% vs. 33% and halving the relapse rate to 19% compared with 55% (44-54); 2) the introduction of molecularly-targeted therapies: FLT3-ITD-positive patients receive sorafenib or gilteritinib; IDH1/2-mutated AML is treated with ivosidenib, enasidenib or HMPL-306; and emerging agents—including MENIN inhibitors for KMT2A-rearranged or NPM1-mutated disease, BET inhibitors targeting epigenetic readers, and DOT1L or LSD1 inhibitors—are under active clinical investigation (55-59); and 3) the VEN plus hypomethylating agents (HMAs) regimen has drastically transformed the survival prognosis of elderly patients with AML, resulting in striking differences in survival data before and after its adoption (60). Before this regimen (in the era of traditional treatments), the prognosis of elderly AML patients was extremely poor: the median OS with intensive chemotherapy (e.g., DA/CAG regimens) was only 10.7−12.0 months, with a 3-year survival rate of approximately 15%; the median OS with low-intensity therapy (HMA monotherapy, low-dose chemotherapy, etc.) was 6−10 months, and the complete remission (CR) rate was less than 30%; for patients receiving only best supportive care (BSC), the median OS was as short as about 2.3 months, with a 1-year survival rate below 5%. After the VEN+HMA regimen was adopted, survival times improved: single-center studies reported better outcomes, with median OS ranging from 14.2 to 24.0 months, and the 1-year OS rate reached 66.8% in patients treated with the optimized DEC3-VEN regimen (61). Notably, specific subgroups benefitted significantly: the median OS for patients with TP53 mutations (historically about 2 months with traditional therapy) extended to 21.3 months, and patients bridged to haplo-HSCT after VEN+HMA achieved a 2-year OS rate of 78.4% (62-65). Collectively, these innovations have increased the national 5-year OS rate for AML to approximately 24%, with low-risk subgroups exceeding 50% and selected molecularly defined subsets reaching >70% (7,66-68), transforming AML from a uniformly fatal disease into a curable condition for an expanding proportion of Chinese patients.

Two decades ago, survival outcomes for ALL patients in China were significantly lower than the current levels. For pediatric ALL patients, the 5-year OS rate was approximately 50%−60% and was constrained by inconsistent treatment protocols across regions, limited access to advanced risk stratification tools and scarce HSCT resources. Adult ALL patients faced even greater challenges, with 5-year OS rates as low as 15%−25%, largely because of the lack of standardized, pediatric-inspired treatment regimens and insufficient monitoring of minimal residual disease (MRD). Over the past two decades, however, ALL survival in China has improved markedly, with greater gains in children than in adults. Pediatric ALL survival increased from 72% to 86% (nearing Western rates of 88%−91%), driven by nationwide Chinese Children’s Cancer Group (CCCG)-ALL-2008/2015 protocols, MRD-guided risk stratification, and 15% first-line allo-HSCT; the CCCG-ALL-2015 protocol achieved 5-year OS of 90.9%. Adult survival increased modestly from 24% to 38% (with a 53.8% increase in 5-year OS using pediatric-inspired regimens). The “Beijing Protocol” for haplo-HSCT addressed donor scarcity, outperforming chemotherapy (80.9% vs. 51.1% disease-free survival in standard-risk ALL) and human leukocyte antigen (HLA)-matched sibling transplantation (68% vs. 46% 3-year OS in MRD-positive ALL) (69-73). Their next-generation sequencing (NGS)-MRD (10^-6 sensitivity) outperformed flow cytometry, guiding transplant decisions (NGS-MRD-positive patients benefited from allo-HSCT: 77% vs. 21% 3-year OS) (74). For Ph+ ALL post-transplant relapse, the 2-year OS was 40.2% (69). China’s pediatric ALL care now aligns with global standards, while the work of Huang’s team resolved domestic challenges (e.g., donor shortages) and informed international guidelines. CD19-directed CAR-T cell therapy has become China’s standard of care for patients with relapsed/refractory B-ALL (75). Nationwide multicenter data show that single-agent CD19 CAR-T cells (4-1BB construct) induce CR in about 85% of heavily pretreated children and adults, with >90% achieving MRD negativity; at 2 years, leukemia-free survival is about 47%, and OS is about 56% (76). To reduce CD19-negative relapse, sequential or coinfused CD19/CD22 dual-target CAR-T cell programs were pioneered; in a 24-patient post-transplant relapse cohort, the 5-year event-free survival and OS rates reached 50% and 75%, respectively, indicating that half of these ultrahigh-risk patients are now cured (77). Bridging to allo-HSCT is recommended whenever feasible; patients who achieve MRD-negative remission and who proceed to transplant enjoy long-term disease-free survival previously unattainable with chemotherapy alone. Professor Huang’s team also optimized CAR-T cell therapy in clinical practice: they recommended bridging to allo-HSCT whenever feasible, as patients who achieve MRD-negative remission post-CAR-T cell therapy and then proceed to transplant attain long-term disease-free survival, which is unattainable with chemotherapy alone (76).

TKIs have dramatically prolonged life expectancy in patients with CML: the 10-year OS reaches 85%−90%, which is in line with that of the general population. Consequently, the disease has become a chronic, manageable condition, and a substantial minority of patients who achieve a deep molecular response can achieve durable TFR after planned discontinuation (18,22,78). For chronic-phase patients who maintain a deep molecular response (MR4.0+, BCR::ABL1IS≤0.01%) for ≥3 years, the 5-year molecular relapse-free rate is about 50% (rising to 68% with favorable factors such as baseline basophils <3% and low Sokal risk), and a bone marrow morphology deep learning model identified high granulocyte maturation (at diagnosis) as a low-cost TFR candidate biomarker (78). The phase III ASC4FIRST trial advanced the use of the STAMP inhibitor asciminib to first-line use. Relative to imatinib/second-generation TKIs, it boosts 24-week MR3.0 (BCR::ABL1IS≤0.1%) by 17 percentage points (vascular events/pulmonary hypertension <1%, 96-week progression-free survival 97%) (79), leading to the 2025 European LeukemiaNet (ELN) Guidelines to list it as a “preferred first-line option” for patients with atherosclerosis/pulmonary vascular high-risk factors (78). For refractory/advanced cases, asciminib + ponatinib (a dual-target regimen) achieves a 12-month major molecular response (MMR, BCR::ABL1IS≤1%) in 76% of patients with ≥2 failed TKIs with BCR::ABL1 gatekeeper mutations (excluding T315I) (no new resistance mutations over 30 months of follow-up) (78). China’s novel third-generation TKI olverembatinib (for T315I-mutated CML) demonstrated 56% of 12-month MMR, 72% of complete cytogenetic response (CCyR), 86% of 3-year OS (36 months of follow-up) and no grade ≥3 cardiovascular events in early first-line trials, prompting its inclusion in China’s 2025 National Reimbursement Drug List, making it the first choice for T315I-mutated patients (78); blast-phase (BP) CML is managed with “third-generation TKI + hypomethylating agent” induction followed by allo-HSCT and 3−5 years of post-transplant TKI maintenance, increasing the 3-year OS from 25% to 55% (78).

Two decades ago, the survival outlook for CLL in China was bleak, plagued by limited diagnostic capacity, scarce effective treatment options, and a lack of systematic survival data. At that time, CLL was not only 10- to 20-fold rarer in China than in Western countries but also often misdiagnosed because of insufficient awareness and immature flow cytometry-based diagnostic protocols. Given the small number of confirmed patients, treatment was largely limited to conventional chemotherapy (e.g., chlorambucil monotherapy) with poor efficacy—5-year OS rates were estimated to be less than 30%, and the median OS rarely exceeded 5 years. Additionally, the absence of targeted therapies and limited access to HSCT left most patients with advanced disease facing an unfavorable prognosis, far behind Western countries where early-stage CLL management had already begun to improve survival. Today, however, population-based and real-world studies have shown clear survival increases since oral targeted agents have become available. Although CLL remains 10- to 20-fold rarer in China than in Western countries, the largest national cohort (Tianjin Chinese Academy of Medical Sciences, n=2,996) recorded a median OS of 9.7 years from treatment initiation and a 5-year survival rate of 70.5%, which is significantly higher than contemporaneous real-world data from the US (24). This advantage is seen across all first-line modalities (chlorambucil, FCR/BR, or BTK/BCL-2 inhibitors) and is most pronounced in older patients and those with mutated-IGHV or intact TP53; younger or TP53-deleted individuals still fare comparably to Western figures, emphasizing the need for broader front-line use of novel agents. The dramatic improvement in survival is closely tied to the widespread adoption of targeted therapies. With ibrutinib, zanubrutinib, and orelabrutinib now reimbursed nationwide, and VEN-based combinations being expanded through clinical protocols, CLL has transitioned from a poor-prognosis disease to a chronic, controllable condition for most Chinese patients. Looking ahead, ongoing fixed-duration BTKi±VEN trials aim to close the remaining outcome gap in ultrahigh-risk subgroups (e.g., TP53-deleted patients), further aligning China’s CLL survival outcomes with leading global standards (80).

Despite survival gains, leukemia remains a major source of disease burden. In 2021, the cluster of leukemia subtypes was responsible for 2.45 million DALYs, ranking fifth among all malignant neoplasms in China (GBD-Cancer-Expenditures 2023) (81).

Risk factors and attributable burden

In 2021, smoking remained the single most important modifiable contributor to leukemia mortality in China, accounting for 53.8% of all leukemia deaths (95% uncertainty interval: 49.3%−58.1%). This translates into approximately 33,000 avoidable deaths annually and reflects both the continued high prevalence of tobacco use among Chinese males (about 52%) and the prolonged latency between smoking initiation and leukemogenesis. High body mass index, the second fastest-growing risk factor, was responsible for 30.8% of deaths, with its PAF rising at +3.1% per year between 2000 and 2021 as China’s age-standardized obesity rate more than doubled. Conversely, the smoking-related PAF decreased by 1.8% per year over the same period, mirroring modest but real successes in tobacco control (28). Occupational exposures, although smaller in absolute terms, still matter. Benzene-attributable deaths accounted for 3.4% of the national total in 2021, but the PAF was 4.7% in males vs. only 0.9% in females; Northeast China and North China, where heavy industry and petro-chemical plants cluster, recorded the highest regional benzene PAF of 6.1%. Together, formaldehyde and residential radon further increased the number of deaths by 3.1%, underscoring the multifactorial nature of environmental risk (29). Under the 2021 exposure profile, these five modifiable factors (smoking, high BMI, occupational benzene, occupational formaldehyde and residential radon) collectively explained 90.7% of all leukemia deaths; if 2030 global non-communicable diseases (NCD) targets (30% relative reduction in smoking prevalence, zero increase in obesity) were achieved, the projected avoidable mortality would reach 21,500 deaths per year, which is equivalent to a 14% reduction in the national leukemia mortality rate.

Summary and prospects

Over the past two decades, China has witnessed a profound epidemiological and clinical transformation in leukemia. The disease burden is no longer driven solely by incidence but is increasingly driven by survival disparities and long-term sequelae. While national incidence rates have plateaued for most subtypes, demographic aging and environmental exposures continue to fuel absolute case growth, particularly for acute leukemias. Notably, survival gains have been substantial but uneven: pediatric ALL and APL patients now approach international benchmarks, whereas adult AML and ALL patients still lag, especially in rural and low-income regions (33). These findings underscore a critical transition—from “how to treat” to “how to deliver precision therapy equitably”.

With respect to prevention, most leukemia prevention efforts in China focus on classic environmental factors (tobacco, benzene, and obesity) (7,36). Future directions should pivot toward molecular epidemiology and early interception. Integrating germline susceptibility panels (e.g., TP53, CEBPA, and DDX41) with large-scale biobanks will allow the identification of high-risk cohorts decades before onset. Parallel efforts should explore the leukemogenic trajectory of CHIP in the Chinese population—its interaction with PM2.5, dietary micronutrients and epigenetic aging clocks—and test whether lifestyle or pharmacologic (e.g., metformin, aspirin) interventions can delay or prevent malignant evolution (7,36). Workplace “omics” studies that couple air-monitoring sensors with real-time epigenomic profiling of exposed workers could provide causal evidence for policy reinforcement.

For therapy, haplo-HSCT has solved the shortage of donors, but its full potential is only beginning to be realized (82-87). Future trial designs should move beyond “transplant vs. chemotherapy” to include dynamic, biology-driven sequences. For AML, induction with VEN plus HMAs followed by MRD-triggered haplo-HSCT, post-transplant maintenance with FLT3 or IDH1/2 inhibitors, and preemptive donor-derived CAR T-cell or natural killer (NK)-cell therapy for emerging MRD constitute an example of an integrated pipeline that needs formal testing. In ALL, dual-target (CD19/CD22 or CD19/CD20) CAR-T cells produced within 72 h (“fast CAR”) could be used as a bridge to haplo-HSCT in high-risk first remission, with transplantation acting as a platform for subsequent CAR-T cell therapy or bispecific T cell engagers boosting. CRISPR-engineered universal CAR NK cells offer an off-the-shelf option to abrogate post-transplant viral infections and leukemic relapse simultaneously. Importantly, such trials must embed health economic evaluations to secure widespread reimbursement.

CML and CLL are becoming “cancers without a cancer ward”. However, the transition to bona fide chronic disease management faces unique challenges. First, cardiovascular- and infection-related mortality now exceed leukemia-specific death; therefore, cardio-oncology clinics integrating NT-proBNP, troponin and cardiac magnetic resonance imaging into routine TKI follow-up are warranted. Second, TFR should be regarded as a “measurable outcome” within the national NCD surveillance network, analogous to HbA1c in diabetes patients. A cloud-based BCR-ABL1 and MRD dashboard, linked to primary care electronic records, could trigger automatic drug reinitiation and prevent TFR failures. Third, for CLL, fixed-duration BTKi-VEN combinations should be tested not only for efficacy but also for the preservation of humoral immunity and quality-adjusted life-years, ensuring that prolonged disease control does not convert into disproportionate infectious morbidity.

Finally, future leukemia control must embrace data equity. Federated learning platforms that allow peripheral hospitals to contribute encrypted molecular and imaging data to central algorithms will circumvent privacy barriers while democratizing precision oncology. Community-based “leukemia interception units” staffed by nurse practitioners armed with artificial intelligence-guided decision support could triage high-risk CHIP carriers and monitor TKI adherence, thereby narrowing the rural-urban survival gap.

Conclusions

The next decade offers a realistic prospect of converting most leukemias in China into either curable or well-controlled chronic conditions. This vision necessitates a unified framework that synthesizes mechanism-based prevention, algorithm-enabled early detection, and adaptive therapeutic sequences (integrating transplantation, targeted agents, and cellular immunotherapy) within an equitable, data-integrated healthcare ecosystem.

Funding Statement

This work was supported by the National Key Research and Development Program of China (No. 2022YFA1103300), the Major Program of the National Natural Science Foundation of China (No. 82293630), the Beijing Municipal Science & Technology Commission (No. Z211100002921071), the Peking University Medicine Fund for the World’s Leading Discipline or Discipline Cluster Development (No. 71003Y3035), the National Natural Science Foundation of China (No. 82570262 and No. 82570208) and the Beijing Natural Science Foundation (No. L252196).

Author contributions

Conceptualization: XJ Huang; Formal analysis: LJ Hu, Y Xia; Funding acquisition: XJ Huang, LJ Hu; Investigation: LJ Hu; Project administration: LJ Hu; Writing – original draft: LJ Hu, Y Xia; Writing – review & editing: LJ Hu, XJ Huang.