Epidemiology of Acute Myeloid Leukemia (AML)

Acute myeloid leukemia (AML) is a common type of leukemia that affects myeloid stem cells that are made into granulocytes, monocytes, and reticulocytes.1 Treatment has rarely changed in the 20th to 21st century, and many minority communities are undertreated.12 We aimed to estimate the prevalence of AML using the Surveillance, Epidemiology, and End Results (SEER) database, a recently launched initiative by the Surveillance Research Program in National Cancer Institute's Division of Cancer and Population Sciences,3 to effect better patient care that is tailored to sociodemographic factors. This study was deemed Institutional Review Board exempt due to its retrospective nature. We performed a cross-sectional analysis of the SEER database by identifying patients with a diagnosis of AML using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 205.0 and ICD-10-CM code C92.00. Electronic medical records of each patient with AML were then analyzed to collect data on age, sex, and self-identified race. SEER provided the overall prevalence of AML based on census data and estimations on populations lost to follow-up. As of 2020, the SEER database has enrolled 43,926,825.00 cases, shown in Table 1. We used the latest available data, which encompasses 2017 to 2018 census data, and identified 9,572.8 with AML, representing an overall prevalence of 0.04%. The prevalence was highest in the 70 to 74 age group, increasing with age. Prevalence in specific racial groups included 0.01% in white, 0.01% in black, 0.01% in American Indian/Alaska Native, and 0.01% in Asian or Pacific Islander patients. Furthermore, the SEER database is 71% white, 12% black, American-Indian 2%, and 15% Asian-Pacific Islanders,4 while the United States population is 76% white, 14% black, 1% American-Indian, and 7% Asian.5 Thus, our prevalence calculation of AML is an underestimation of the white and black population, and an overestimation of the Hispanic and Asian population. It is also likely that there are more patients that are unaccounted for due to U.S. residency status, health care availability, and limitations of the census. Chi-squared test of independence show that there is no significant difference between the SEER and the U.S. Census population. There was not a significant relationship between the two populations, chi-square (3, N = 304,167,848) = 3,404,209.8855, p < 0.00001 (Table 2). This result suggests that these populations are statistically similar and allow an estimation of the U.S. population using SEER. Altogether, our data suggest AML is a common leukemia across all racial groups. Further epidemiologic studies that are not restricted by billing codes may validate our findings. Understanding sociodemographic factors can increase clinical practice diversity and inclusivity to increase diagnosis and treatment among minority populations.