Rates and Trends of Pediatric Acute Lymphoblastic Leukemia — United States, 2001–2014

Acute lymphoblastic leukemia (ALL) is the most prevalent cancer among children and adolescents in the United States, representing 20% of all cancers diagnosed in persons aged <20 years, or >3,000 new cases each year (1). Past studies reported increasing trends of ALL overall and among Hispanics, but these represented ≤28% of the U.S. population and did not provide state-based estimates (1-3). To describe U.S. ALL incidence rates and trends among persons aged <20 years during 2001-2014, CDC analyzed rigorous data (based on established publication criteria) from the United States Cancer Statistics data set, which includes incidence data on approximately 15,000 new cases per year of all types of invasive cancer among children and adolescents aged <20 years (4). The data set represented 98% of the U.S. population during the study period. Overall incidence of pediatric ALL during 2001-2014 was 34.0 cases per 1 million persons and among all racial/ethnic groups was highest among Hispanics (42.9 per 1 million). Both overall and among Hispanics, pediatric ALL incidence increased during 2001-2008 and remained stable during 2008-2014. ALL incidence was higher in the West than in any other U.S. Census region. State-specific data indicated that the highest rates of pediatric ALL incidence were in California, New Mexico, and Vermont. These demographic and geographic ALL incidence data might better inform public health interventions targeting the following areas: exposures to recognized risk factors for leukemia; ALL treatment, including clinical trial enrollment; survivorship care planning; and studies designed to understand the factors affecting changes in pediatric cancer incidence.

Acute lymphoblastic leukemia (ALL) is the most prevalent cancer among children and adolescents in the United States, representing 20% of all cancers diagnosed in persons aged <20 years, or >3,000 new cases each year (1). Past studies reported increasing trends of ALL overall and among Hispanics, but these represented ≤28% of the U.S. population and did not provide state-based estimates (1)(2)(3). To describe U.S. ALL incidence rates and trends among persons aged <20 years during 2001-2014, CDC analyzed rigorous data (based on established publication criteria) from the United States Cancer Statistics data set, which includes incidence data on approximately 15,000 new cases per year of all types of invasive cancer among children and adolescents aged <20 years (4). The data set represented 98% of the U.S. population during the study period. Overall incidence of pediatric ALL during 2001-2014 was 34.0 cases per 1 million persons and among all racial/ethnic groups was highest among Hispanics (42.9 per 1 million). Both overall and among Hispanics, pediatric ALL incidence increased during 2001-2008 and remained stable during 2008-2014. ALL incidence was higher in the West than in any other U.S. Census region. Statespecific data indicated that the highest rates of pediatric ALL incidence were in California, New Mexico, and Vermont. These demographic and geographic ALL incidence data might better inform public health interventions targeting the following areas: exposures to recognized risk factors for leukemia; ALL treatment, including clinical trial enrollment; survivorship care planning; and studies designed to understand the factors affecting changes in pediatric cancer incidence.
The United States Cancer Statistics data set includes cancer incidence data from CDC's National Program of Cancer Registries and the National Cancer Institute's Surveillance, Epidemiology, and End Results program (4 Age-adjusted rates were calculated using statistical software. All rates were expressed per 1 million persons and were ageadjusted to the 2000 U.S. standard population. ¶ Age-adjusted incidence trends were quantified using annual percent change (APC) calculated using joinpoint regression. Statistically significant APCs were different from zero (p<0.05). A maximum of two joinpoints were used to determine a change of direction in trends during the study period. Rates and trends were estimated by sex, age group, race/ethnicity, state, U.S. Census region,** county-level economic status, and rural/ urban status.
During 2001-2014, a total of 38,136 new pediatric ALL cases were diagnosed in the United States (Table). Overall incidence of ALL was 34.0 cases per 1 million. Rates were highest in children aged 1-4 years (75.2 per 1 million) and were higher in males (38.0) than in females (29.7). Among all racial/ ethnic groups, the highest incidence rate (42.9 per 1 million) was among Hispanics, followed by non-Hispanic whites (34.2 per 1 million). The lowest incidence (18.7) occurred among non-Hispanic blacks. Pediatric ALL incidence rates in the 25% of U.S. counties with the highest economic status were higher than rates in the 25% of counties with the lowest economic status and were higher in metropolitan areas with ≥1 million persons than in nonmetropolitan areas. Rates were highest in the West (38.5) followed by the Northeast (34. 8

TABLE. Age-adjusted incidence* of acute lymphoblastic leukemia † in persons aged <20 years and annual percentage change (APC) in rates, by selected characteristics -United States, § 2001-2014 Characteristic
No.

Discussion
Consistent with other published data, this analysis found that rates of ALL were highest in males, children aged 1-4 years, and Hispanics (1). Rates varied by state and region and were highest in the West U.S. Census Region. This report, using more recent data with broader population coverage than past studies (1-3), confirms an increase in pediatric ALL overall and among Hispanics (2001-2008) and also documents a subsequent period of stable trends overall and among Hispanics (2008-2014).
High rates of pediatric ALL in the Hispanic population might explain high ALL rates in the West U.S. Census Region and in other specific states, given the high proportion of Hispanics in many of these areas. † † Past studies documenting increasing incidence of pediatric ALL in Hispanics focused † † https://www.census.gov/prod/cen2010/briefs/c2010br-04.pdf. on earlier periods using the Surveillance, Epidemiology, and End Results database or the California Cancer Registry (2,3). Recent stable trends in ALL rates in Hispanic populations (2008-2014) might indicate a change after 2 decades of documented increasing trends. The cause for the higher rates of ALL in Hispanic populations and the increase during 2001-2008 is unknown; however, past studies have evaluated such factors as genetic susceptibility, disproportionate environmental exposure to household chemicals, or racial and ethnic disparities in parents' exposures to chemicals at work (2,3). Other studies have hypothesized that the increasing trends in obesity among Hispanics might explain the increasing trends of ALL incidence among this population (2).
This report documents higher rates of ALL in persons aged <20 years living in counties in metropolitan areas with ≥1 million population and in counties in the top 25th income percentile. Past studies of pediatric leukemia have investigated possible associations with higher economic status or increased exposure to air pollution that is often found in large metropolitan areas (5,6). Etiologic studies examining potential causes of pediatric leukemia have documented associations between leukemia and exposures to solvents, traffic, pesticides, tobacco smoke, or radiation, or to specific nutritional exposures (7).
The findings in this report are subject to at least five limitations. First, the District of Columbia, Mississippi, and Nevada were excluded because of incomplete trend data, which limits the representativeness of the results. Second, although the United States Cancer Statistics data publication standards yield high quality data, misclassifications of race and ethnicity exist and might underestimate rates in American Indians, Alaska Natives, and Hispanics (8); ongoing procedures are used to ensure that this information is as accurate as possible. Third, the U.S. Census population estimates used in rate denominators might undercount some groups, including children and Hispanics, which could artificially raise incidence rates (3). Fourth, improvement in case ascertainment through advancements in electronic pathology reporting might affect trends: rates might appear to increase because current cancer registration methods are more accurately recording cases that were previously under-recorded. Finally, the possibility of a statistical error exists when analyzing subgroups with small numbers. Although APCs that are close to the significance cutoff might be truly significant, future studies will be needed to validate and monitor trends.
These recent state-based demographic cancer data can help local and national cancer control programs assess needs, allocate resources, and guide policy and public health strategies that can reduce cancer risk and improve the care of children and adolescents with ALL. Because cancer clinical trial participation has become an increasingly important part of quality clinical

Summary
What is already known about this topic? Acute lymphoblastic leukemia (ALL) is the most common cancer in children and adolescents in the United States. Past studies using ≤28% population coverage have described increasing incidence of pediatric ALL, especially in Hispanic populations.
What is added by this report?
Analysis of data covering 98% of the U.S. population indicated that the incidence of pediatric ALL increased during 2001-2008 overall and for Hispanics, but then was stable during 2008-2014. The cause for the higher rates of ALL in Hispanic populations and the increase during 2001-2008 is unknown. Incidence of pediatric leukemia during 2001-2014 was highest in the West U.S. Census Region, possibly reflecting the high proportion of Hispanics in many of the region's constituent states.
What are the implications for public health practice?
Increasing incidence of pediatric ALL in certain demographic groups might necessitate changes to cancer control planning, affecting treatment and survivorship care. Continued cancer surveillance will be important in guiding future research, including etiologic studies.
care (9), many state health departments have created cancer control plans that aim to address the economic and sociocultural barriers that limit certain groups enrolling in these trials. § § Knowledge about rates and trends of pediatric ALL might help tailor the goals of these programs to address local and disease-specific needs. In addition, as incidence and survival of pediatric ALL increase (1), public health professionals can use recent ALL incidence data to improve local cancer survivorship programs that address chronic disease management, screen for late effects, and provide resources to help patients maintain a high quality of life (10). Public health planners can prioritize issues pertinent to pediatric cancer survivors such as transitioning to adult care and accessing the educational resources that might be available to these patients. Finally, health care professionals and researchers can use surveillance data to inform research questions. Continued surveillance data will be needed to further track incidence changes and public health needs relative to specific demographic groups and geographic areas. § § https://www.cdc.gov/cancer/ncccp/index.htm.