Abstract
PURPOSE: Advances in the treatment of childhood hematologic malignancies have led to improvements in survival in several of these conditions during the past few decades. Here we use the novel method of modeled period analysis to disclose most recent trends in survival among patients diagnosed with hematologic malignancies at less than 15 years of age1.
METHODS: We estimated trends in 5-, 10-, and 25-year survival in children with common hematologic malignancies in the US over three 5-year intervals: 1990–94, 1995–99, and 2000–04, using data from the Surveillance, Epidemiology, and End Results program. Expected survival for 2005–09 was estimated by modeling from trends in the preceding intervals2.
RESULTS: Major improvements in 5- and 10-year relative survival between 1990–94 and 2000–04 were seen in acute lymphoblastic leukemia (ALL) (+7.3 and +10.4% points, respectively), acute non-lymphoblastic leukemia (ANLL) (+18.0 and +20.4% points) and non-Hodgkin lymphoma (NHL) (+11.1 and +13.9% points). Improvements in survival for patients with ALL were particularly strong in older patients, so that the significance of age as a prognostic factor decreased for children diagnosed with ALL in the early 21st century. For Hodgkin’s disease (HD), 5 and 10-year survival were as high as 96.1% and 94.1% in 1990–94 and did not show recent changes. However, it is notable that the 5- and 10-year survivals for HD in 2000–04 were identical, that is, no further deaths occurred between 5 and 10 years. Twenty-five year survival was lower for all of the malignancies examined, with 25-year survivals of about 80% for ALL, HD, and NHL and about 50% for ANLL. Projected 10-year survivals for children diagnosed in 2005–09 are 88.0%, 63.9%, 90.6%, and 94.3% for ALL, ANLL, NHL, and HD, respectively.
CONCLUSION: Application of period analysis to a population-based study of hematologic malignancies reveals ongoing strong improvement in 3 of the 4 common childhood hematologic malignancies. Projections for 2005–09 suggest chances of 10-year survival to be close to 90% or higher for children with ALL, HD, or NHL and to approach 65% for children with ANLL. Long term outcomes remain problematic, with some excess loss of life occurring as late as 25 years after diagnosis.
Period estimates of 10-year survival for childhood hematologic malignancies
| Malignancy . | 1990–94 (SE) . | 1995–99 (SE) . | 2000–04 (SE) . | Difference . | P-val . | Projected for 2005–09 (SE) . |
|---|---|---|---|---|---|---|
| ALL-total | 73.4 (1.5) | 81.0 (1.3) | 83.8 (1.2) | +10.4 | <0.0001 | 88.0 (1.4) |
| ALL-0-2 yr | 72.3 (3.3) | 80.9 (2.9) | 83.4 (2.8) | +11.1 | 0.01 | 87.9 (3.0) |
| ALL-3–4 yr | 80.9 (2.4) | 84.9 (2.1) | 88.9 (1.8) | +8.0 | 0.006 | 91.6 (2.0) |
| ALL-5–9 yr | 71.4 (2.7) | 80.4 (2.4) | 81.4 (2.2) | +10.0 | 0.004 | 86.1 (2.6) |
| ALL-10–14 yr | 61.3 (2.1) | 72.3 (4.1) | 78.2 (3.6) | +16.9 | 0.007 | 83.9 (4.0) |
| ANLL | 38.7 (4.0) | 42.7 (3.8) | 59.1 (3.7) | +20.4 | 0.0005 | 63.9 (4.8) |
| HD | 94.4 (2.1) | 90.8 (2.6) | 95.4 (1.8) | +1.0 | 0.79 | 94.3 (3.0) |
| NHL | 73.0 (4.0) | 81.7 (3.4) | 86.9 (2.6) | +13.9 | 0.005 | 90.6 (2.8) |
| Malignancy . | 1990–94 (SE) . | 1995–99 (SE) . | 2000–04 (SE) . | Difference . | P-val . | Projected for 2005–09 (SE) . |
|---|---|---|---|---|---|---|
| ALL-total | 73.4 (1.5) | 81.0 (1.3) | 83.8 (1.2) | +10.4 | <0.0001 | 88.0 (1.4) |
| ALL-0-2 yr | 72.3 (3.3) | 80.9 (2.9) | 83.4 (2.8) | +11.1 | 0.01 | 87.9 (3.0) |
| ALL-3–4 yr | 80.9 (2.4) | 84.9 (2.1) | 88.9 (1.8) | +8.0 | 0.006 | 91.6 (2.0) |
| ALL-5–9 yr | 71.4 (2.7) | 80.4 (2.4) | 81.4 (2.2) | +10.0 | 0.004 | 86.1 (2.6) |
| ALL-10–14 yr | 61.3 (2.1) | 72.3 (4.1) | 78.2 (3.6) | +16.9 | 0.007 | 83.9 (4.0) |
| ANLL | 38.7 (4.0) | 42.7 (3.8) | 59.1 (3.7) | +20.4 | 0.0005 | 63.9 (4.8) |
| HD | 94.4 (2.1) | 90.8 (2.6) | 95.4 (1.8) | +1.0 | 0.79 | 94.3 (3.0) |
| NHL | 73.0 (4.0) | 81.7 (3.4) | 86.9 (2.6) | +13.9 | 0.005 | 90.6 (2.8) |
References
Author notes
Disclosure: No relevant conflicts of interest to declare.
