Trimodal age‐specific incidence patterns for Burkitt lymphoma in the United States, 1973–2005 |
| |
Authors: | Sam M. Mbulaiteye William F. Anderson Kishor Bhatia Philip S. Rosenberg Martha S. Linet Susan S. Devesa |
| |
Affiliation: | 1. Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics (DCEG), National Cancer Institute, Bethesda, MDTel: +1‐301‐496‐8115;2. Fax: +1‐301‐402‐0817;3. Biostatistics Branch, DCEG, National Cancer Institute, Bethesda, MD;4. Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics (DCEG), National Cancer Institute, Bethesda, MD;5. Radiation Epidemiology Branch, DCEG, National Cancer Institute, Bethesda, MD |
| |
Abstract: | Burkitt lymphoma (BL) is a unique B‐cell non‐Hodgkin lymphoma with 3 established clinical‐epidemiological variants: endemic, sporadic and AIDS‐related BL. BL variants show characteristic dysregulation of MYC gene, but the causes of MYC dysregulation or BL arising at different ages are poorly understood. Therefore, we examined population‐based BL incidence patterns in the United States to determine age‐related risk. BL case and population data were obtained from the NCI's Surveillance, Epidemiology and End Results Databases (1973–2005). Standard cross‐sectional age‐standardized and age‐specific incidence rates were stratified by sex and race and supplemented with age–period–cohort models. We analyzed 3,058 BL cases diagnosed during 1,160,300,297 person‐years of observation. Age‐standardized incidence rates rose 6.8% per year (95% CI 4.5–9.1) for males and 7.1% (95% CI 3.2–11.1) for females during the study period. The rate among males was 3.2 times that among females, and among Whites 1.3 times that among Blacks. Male‐to‐female incidence rate ratios did not differ by race, but were 4.2 for pediatric (0–19 years), 4.1 for adult (20–59 years) and 2.0 for geriatric (≥60 years) BL. Cross‐sectional age‐specific rates showed 2 separate peaks among males and females, near ages 10 and 75 years, and a 3rd peak near age 40 years among males. The tri/bimodal incidence pattern was present in sensitivity analyses excluding registries with many HIV/AIDS cases and in period‐specific, cohort‐specific analyses. To our knowledge, tri/bimodal incidence patterns have not previously been reported for BL. Trimodal/bimodal BL suggests heterogeneity in etiology or biology of BL diagnosed at different ages in males and females. © 2009 UICC. |
| |
Keywords: | non‐Hodgkin lymphoma epidemiology multistage cancer models tri/bimodal cancer Epstein‐Barr virus |
|
|