Blood Cancers: Survival Up for Every Major Type, but Diagnosis Often Comes Late

Leukaemia, lymphoma and myeloma together accounted for an estimated 1.34 million new cancer diagnoses and 666,000 deaths worldwide in 2024. They are not one disease but dozens. For some types, five-year survival in the United States has risen fourfold or more since the 1970s; for others, the gains have been much smaller, and, where registries have looked closely, many people still wait months for a diagnosis or are diagnosed after an emergency admission.

Blood cancers reach people at every age, from young children to people in their eighties, and through very different routes: an urgent admission with acute leukaemia, months of vague symptoms before myeloma is found, or a routine blood test that turns up a chronic leukaemia needing no treatment for years. This article keeps each number attached to the specific type it describes, and says so where types are grouped. It draws on the global estimates, the long-running US registries and, as a worked case, a UK population-based registry with unusually detailed data on every blood cancer diagnosed in its area. None of it predicts what will happen to any individual.

Taken together, blood cancers rank among the most common cancers in the world

GLOBOCAN 2024, the International Agency for Research on Cancer's latest estimates for 186 countries, reports blood cancers in four groups.^[1] Non-Hodgkin lymphoma was the largest, with an estimated 562,793 new cases, followed by leukaemia (495,113), multiple myeloma (196,157) and Hodgkin lymphoma (84,555). Adding the four together gives about 1.34 million new cases, around 6.5% of the 20.6 million new cancers estimated worldwide, and about 666,000 deaths, around 6.8% of all cancer deaths. That sum is ours, not IARC's, and it is an undercount of blood cancer as a whole: GLOBOCAN's four groups do not separately cover conditions such as myelodysplastic syndromes and myeloproliferative neoplasms, which are classed as blood cancers too.

Source: Sung et al., CA: A Cancer Journal for Clinicians 2026, Table 1 (both sexes).^[1] These are modelled estimates; in many countries without full registry coverage they rest on limited local data.

Even these groups hide very different diseases. "Leukaemia" in GLOBOCAN combines acute leukaemias, which develop over days or weeks, with chronic leukaemias that may be watched for years before any treatment. That is why population registries that record each subtype matter. The UK's Haematological Malignancy Research Network (HMRN) is one: it records the blood cancers diagnosed in a defined UK population of about 3.6 million people (nearly 4 million in later reports), subtype by subtype, and follows people over time.^[4] It recorded 10,729 haematological malignancies over five years from 2004, with a median age at diagnosis of 70.6, and found that for several myeloid and lymphoid subtypes the rate in males was more than double that in females. It found no association between incidence and area deprivation, suggesting age and sex matter more for who is diagnosed.^[4] Its later reports describe more than 20 myeloid subtypes and more than 20 lymphoma subtypes separately.^[5][6]

Most of these figures describe adults. Children's blood cancers, acute lymphoblastic leukaemia above all, have their own evidence base and are covered in our article on children's and young people's cancer.

Survival has risen for every major type, but from very different starting points

The US Surveillance, Epidemiology, and End Results (SEER) programme has tracked survival in the same registries since 1975. Comparing people diagnosed in 1975 with those diagnosed in 2018, five-year relative survival (SEER's single-year values, not its modelled trend) rose for every major blood cancer type shown below.^{[2][7][8][9][10][11]} The largest proportional gains were for acute myeloid leukaemia (AML), from 5.5% to 33.2%, and chronic myeloid leukaemia (CML), from 18.0% to 72.4%. Myeloma rose from 26.1% to 65.9%. Hodgkin lymphoma and chronic lymphocytic leukaemia (CLL) started higher and reached about 94% and 93%. Single-year values move from year to year, so the direction is more certain than the exact size of each change. Our guide to survival statistics explains these measures.

Death rates point the same way. Over 2015 to 2024, age-adjusted US death rates fell on average by 3.1% a year for Hodgkin lymphoma, 3.0% for myeloma, 2.4% for non-Hodgkin lymphoma, 1.9% for CLL, 0.9% for AML and 0.7% for CML.^{[11][8][9][10][7][2]} Falling mortality alongside rising survival is a stronger sign of real progress than survival gains alone, which can partly reflect earlier detection.

The contrast between types remains stark. For people diagnosed in 2016 to 2022 in the wider SEER 21 registries, five-year relative survival was 90.2% for CLL and 89.3% for Hodgkin lymphoma, but 63.7% for myeloma and 33.4% for AML.^{[10][11][8][7]} HMRN's UK data show the same spread within myeloid cancers alone: among 5,231 myeloid diagnoses in 2004 to 2015, five-year relative survival ranged from 2.6% for therapy-related AML (an estimate from small numbers, with a 95% confidence interval of 0.4% to 9.0%) to 89.8% for CML.^[5] For all myeloid subtypes combined it was 48.8% in males and 60.4% in females. Among 5,796 lymphomas diagnosed in 2004 to 2014, HMRN found that people with nodular lymphocyte-predominant Hodgkin lymphoma had survival approaching that of the general population, while fewer than a third of people with some other B-cell lymphomas (mantle cell lymphoma, for example) and T-cell lymphomas (such as peripheral T-cell lymphoma) survived five years or more.^[6]

CML shows how far a single type can move. In Sweden's national cancer registry, 2,662 people were diagnosed with CML between 1973 and 2013; those diagnosed in 2013 were expected to "lose < 3 life-years as a result of CML" on average.^[12] A later Swedish CML register study of 991 adults diagnosed in chronic phase in 2007 to 2017 found the largest loss in any age and sex group was 5.7 years, for women diagnosed at 45; across groups, the loss was 9% to 15% of expected remaining life. Measured in quality-adjusted life years, the loss was 29% to 33%, a reminder that living with a chronic blood cancer and its long-term treatment has costs that survival figures alone do not show.^[13]

Myeloma can take months to diagnose, and many blood cancers are found in an emergency

Blood cancers often cause symptoms, such as tiredness, pain, infections or weight loss, that are common in many other conditions, and some cause none at all. HMRN asked people diagnosed in 2004 to 2011 about the path to their diagnosis; 3,329 returned questionnaires.^[3] Around a third had no symptoms at diagnosis. Among those who did, the median total interval from first symptom to diagnosis was 41 days for AML, 98 days for diffuse large B-cell lymphoma (DLBCL) and 163 days for myeloma. A quarter of people with myeloma waited 306 days or longer. Myeloma was not the slowest to be diagnosed: the study's longest median intervals were for some slow-growing types, 172 days for marginal zone lymphoma and 215 days for myeloproliferative neoplasms. The authors noted that some common symptoms were not in UK referral guidelines, and called for measures such as public education and decision aids for GPs.^[3]

England's 2010 National Cancer Patient Experience Survey found the same pattern from a different angle. Of 1,854 people with multiple myeloma, 50.6% had seen their GP three or more times before being referred to hospital, the highest of the examples given in the study's summary, compared with 41.3% for pancreatic cancer, 10.1% for melanoma and 7.4% for breast cancer.^[14] Repeated visits reflect how hard myeloma is to recognise from its early signs; they are a gap in the diagnostic pathway, not a failure by the person. Our article on the route from symptom to diagnosis sets these figures beside other cancers.

When diagnosis is not reached through planned routes, it often comes through an emergency. HMRN's studies show this for several types:

• Myeloma: for 441 people diagnosed in 2012 to 2013, emergency presentation was the most common single route to diagnosis, at 28.1%, ahead of urgent GP referral (19.0%) and the two-week-wait cancer referral (17.2%). People diagnosed as emergencies were the most likely to have two or more of the CRAB features (raised calcium, kidney failure, anaemia, bone lesions) used to assess myeloma, and had significantly worse survival.^[15]
• Diffuse large B-cell lymphoma: two in every five of 1,660 adults diagnosed in 2004 to 2011 were diagnosed after an emergency admission. Among those given curative-intent chemotherapy, the rate of death in the first month was four times that of people diagnosed by other routes (adjusted hazard ratio 4.0, 95% CI 1.9 to 8.2), but there was no lasting difference for people who survived 12 months or more.^[16]
• Classical Hodgkin lymphoma: 18.7% of 971 people diagnosed in 2004 to 2015 were admitted as emergencies before diagnosis. They tended to have more advanced disease, and three-year relative survival was 68.4%, against 89.8% for people diagnosed by other routes; after allowing for prognostic factors, the difference was no longer seen.^[17]

Where people live, and who they are, still shapes the odds

The CONCORD-3 programme compared survival for 37.5 million people diagnosed with cancer in 2000 to 2014, using 322 population-based registries in 71 countries and territories.^[18] The authors noted that survival from lymphoid and myeloid cancers in South Korea, Taiwan and China was generally lower than elsewhere: five-year net survival from lymphoid malignancies was 52.5%, 50.5% and 38.3% respectively, and from myeloid malignancies 45.9%, 33.4% and 24.8% (the abstract gives the three figures in that order; the study covers people diagnosed up to 2014).^[18] For children diagnosed in 2010 to 2014, five-year net survival from acute lymphoblastic leukaemia ranged from 49.8% in Ecuador to 95.2% in Finland. Our article on survival compared across countries sets out what drives gaps like these.

Part of the gap is likely to be access to diagnosis and medicines. Imatinib is one of the tyrosine kinase inhibitor drugs used to treat CML. Through a drug-donation programme, the Glivec International Patient Assistance Program, 33,985 people with CML in 94 low- and middle-income countries received imatinib. Their three-year overall survival was 89.4% (95% CI 88.9 to 89.9).^[19] In 2002, 85.2% of people in the programme had waited more than a year between diagnosis and treatment, falling to 15.5% by 2010; the study does not say why, and one likely reason is that people reaching a new programme in its early years had often been diagnosed long before. The mean age at diagnosis was 38.5 years, much younger than the US median age at diagnosis of 67 (a median for 2019 to 2023, so not a like-for-like comparison), and the risk of death was higher among people in Latin America and Africa than in other regions.^[19][2] These are people who reached a treatment programme, not whole populations, so the results cannot stand for survival across those countries.

Differences also run within countries. In the US, myeloma incidence among non-Hispanic Black people was 17.9 per 100,000 men and 13.4 per 100,000 women, against 7.4 per 100,000 for the whole population (age-adjusted, 2019 to 2023), and Black Americans also had the highest myeloma death rates.^[8] HMRN found incidence to be strongly patterned by sex and age rather than by deprivation.^[4]

What the evidence asks of us

The figures point to clear priorities for health services, commissioners, funders and governments.

• Shorten the road to diagnosis, especially for myeloma. A median of 163 days from first symptom to a myeloma diagnosis, half of people with myeloma seeing a GP three or more times before referral, and emergency presentation as the most common route all point to a pathway that is not recognising this cancer early enough.^[3][14][15] The HMRN researchers called for public education and decision support for GPs, and for referral guidance that reflects the symptoms people with blood cancers actually report.^[3]
• Treat emergency diagnosis as a measure of the system. Two in five people with DLBCL and more than one in four with myeloma in the HMRN area were diagnosed through an emergency route, with higher early mortality for DLBCL.^[16][15] Publishing route-to-diagnosis figures by subtype would show where people are being missed.
• Make the medicines that changed outcomes reachable everywhere. Survival from CML rose from 18% to 72% in the US over four decades, and people treated through a donation programme in low- and middle-income countries had three-year overall survival of 89.4%.^[2][19] Wide gaps in lymphoid and myeloid cancer survival between countries suggest that access to diagnosis and treatment, not only biology, is at stake.^[18]
• Count blood cancers type by type. Global estimates group dozens of diseases into four categories and leave some out entirely.^[1] Population registries that record each subtype, its route to diagnosis and what happens next, as HMRN does, are what allow gaps to be measured; funders and governments should support that kind of registration.
• Plan for long-term living with blood cancer. As more people live for years with chronic types such as CML and CLL, services need to support quality of life as well as survival: the Swedish data show the proportional loss in quality-adjusted life expectancy (29% to 33%) is far larger than the loss in life expectancy (9% to 15%).^[13]

Language: this piece says "people with cancer" and "people living with and after cancer", and describes the illness and its treatment plainly, without combat metaphors. Clinical terms such as "stage", "incidence" and "net survival" are used in the sense the cited source defines them.

Why we're publishing this

Blood cancer charities and support groups make the case that blood cancers are many different diseases, that they are often hard to spot, and that progress for one type does not mean progress for all. We built this overview, part of our cancer research library, to put the international evidence for that case in one place, checked and sourced, ready to show a commissioner, a funder, a minister or a journalist. The case is theirs. Our part is to set out plainly what the data show: real and hard-won gains in survival, alongside diagnostic delays and gaps in access that services can close.

How to read this data

Global figures are GLOBOCAN modelled estimates of incidence and mortality for 2024; in countries without full registry coverage they rest on limited data, and their four blood cancer groups do not cover every haematological cancer. Our combined total is a simple sum of those four groups. US survival figures are SEER five-year relative survival, which allows for deaths from other causes; values for 1975 and 2018 are single-year figures from the SEER 8 registries (SEER's "observed" data points, not its modelled trend), and recent figures for 2016 to 2022 come from the wider SEER 21 registries, so the two are not strictly comparable. HMRN figures come from one UK region of about 4 million people and from separate studies with different periods; they are a worked example of what detailed registration can show, not national estimates. Diagnostic interval data are self-reported by people who chose to answer a questionnaire. CONCORD-3 figures relate to people diagnosed up to 2014, and the drug-programme figures describe people who reached treatment, not whole populations. Hazard ratios compare risk over time between groups; they are not percentages. Most detailed evidence comes from high-income countries. Where we could see only an abstract, we used only what the abstract says.

Use this data

Free to cite with attribution to Better Health For Everyone. Charities, campaigners and journalists are welcome to reuse these figures and charts. For the underlying figures as a spreadsheet, get in touch.

Important: This article summarises published research for general information. It is not medical advice and cannot be used to diagnose, treat or predict anything about an individual. If you have concerns about your own or someone else's health, please speak to a qualified health professional.

Sources

1. Sung H, Filho AM, Laversanne M, Ferlay J, Siegel RL, Soerjomataram I, Jemal A, Bray F. Global cancer statistics 2024: GLOBOCAN estimates of incidence and mortality worldwide for 34 cancers in 186 countries. CA: A Cancer Journal for Clinicians, 2026. europepmc.org/article/PMC/PMC13343830
2. National Cancer Institute. Cancer Stat Facts: Leukemia — Chronic Myeloid Leukemia (CML). SEER Program, NCI, 2026. seer.cancer.gov/statfacts/html/cmyl.html
3. Howell DA, Smith AG, Jack A, et al. Time-to-diagnosis and symptoms of myeloma, lymphomas and leukaemias: a report from the Haematological Malignancy Research Network. BMC Hematology, 2013. europepmc.org/article/MED/24238148
4. Smith A, Howell D, Patmore R, et al. Incidence of haematological malignancy by sub-type: a report from the Haematological Malignancy Research Network. British Journal of Cancer, 2011. europepmc.org/article/MED/22045184
5. Roman E, Smith A, Appleton S, et al. Myeloid malignancies in the real-world: occurrence, progression and survival in the UK's population-based Haematological Malignancy Research Network 2004-15. Cancer Epidemiology, 2016. europepmc.org/article/MED/27090942
6. Smith A, Crouch S, Lax S, et al. Lymphoma incidence, survival and prevalence 2004-2014: sub-type analyses from the UK's Haematological Malignancy Research Network. British Journal of Cancer, 2015. europepmc.org/article/MED/25867256
7. National Cancer Institute. Cancer Stat Facts: Leukemia — Acute Myeloid Leukemia (AML). SEER Program, NCI, 2026. seer.cancer.gov/statfacts/html/amyl.html
8. National Cancer Institute. Cancer Stat Facts: Myeloma. SEER Program, NCI, 2026. seer.cancer.gov/statfacts/html/mulmy.html
9. National Cancer Institute. Cancer Stat Facts: Non-Hodgkin Lymphoma. SEER Program, NCI, 2026. seer.cancer.gov/statfacts/html/nhl.html
10. National Cancer Institute. Cancer Stat Facts: Leukemia — Chronic Lymphocytic Leukemia (CLL). SEER Program, NCI, 2026. seer.cancer.gov/statfacts/html/clyl.html
11. National Cancer Institute. Cancer Stat Facts: Hodgkin Lymphoma. SEER Program, NCI, 2026. seer.cancer.gov/statfacts/html/hodg.html
12. Bower H, Björkholm M, Dickman PW, et al. Life expectancy of patients with chronic myeloid leukemia approaches the life expectancy of the general population. Journal of Clinical Oncology, 2016. europepmc.org/article/MED/27325849
13. Chen EY, Dahlén T, Stenke L, et al. Loss in overall and quality-adjusted life expectancy for patients with chronic-phase chronic myeloid leukemia. European Journal of Haematology, 2025. europepmc.org/article/MED/39501755
14. Lyratzopoulos G, Neal RD, Barbiere JM, et al. Variation in number of general practitioner consultations before hospital referral for cancer (2010 National Cancer Patient Experience Survey, England). Lancet Oncology, 2012. europepmc.org/article/MED/22365494
15. Howell D, Smith A, Appleton S, et al. Multiple myeloma: routes to diagnosis, clinical characteristics and survival — findings from a UK population-based study. British Journal of Haematology, 2017. europepmc.org/article/MED/28146275
16. Kane E, Howell D, Smith A, et al. Emergency admission and survival from aggressive non-Hodgkin lymphoma: a report from the UK's population-based Haematological Malignancy Research Network. European Journal of Cancer, 2017. europepmc.org/article/MED/28412589
17. Lamb MJ, Roman E, Howell DA, et al. Hodgkin lymphoma detection and survival: findings from the Haematological Malignancy Research Network. BJGP Open, 2019. europepmc.org/article/MED/31822492
18. Allemani C, et al. Global surveillance of trends in cancer survival 2000-14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries. The Lancet, 2018. europepmc.org/article/MED/29395269
19. Mendizabal AM, et al. Regional variations in age at diagnosis and overall survival among patients with chronic myeloid leukemia from low and middle income countries. Cancer Epidemiology, 2013. europepmc.org/article/MED/23411044