Association of Neighborhood Crime with AML Molecular Subtypes and Survival

Introduction: There are well-documented differences in acute myeloid leukemia (AML) survival based on race and ethnicity based on population-level data from Surveillance, Epidemiology, and End Results (SEER) cancer registry and the National Cancer Data Base (NCDB). The inferior survival in non-Hispanic black (NHB) patients is unexpected, given the higher incidence of favorable cytogenetics in NHB compared to non-Hispanic White (NHW) AML patients. A previously published study from our group found that census tract-level variables of disadvantage, affluence, and segregation accounted for nearly all the disparity in AML survival between NHB and NHW (Abraham et al., 2022). Neighborhood social and physical environment can promote cancer development through multiple mechanisms. Neighborhood violence has been shown to induce a strong inflammatory response in healthy individuals (Miller et al. 2022), and recent data highlight the independent negative prognostic role of an upregulated inflammatory signature in AML (Lasry et al. Nature Cancer 2023).

Hypothesis: Exposure to neighborhood crime may act through neurobiological mechanisms to influence leukemia clonal architecture and outcomes.

Methods: The Chicago Area Leukemia Registry includes 822 adult patients in the Chicagoland area diagnosed with Acute Myeloid Leukemia between 2012 and 2018. We initially used data from 216 patients who resided in Chicago, for whom detailed census tract-level crime rates were available at the census tract from the Chicago Department of Public Health. We accessed the most recently available data for the five-year period 2018-2022 on 13 specific groups of related crimes (e.g., aggravated assault/battery, homicide, larceny, property crime, robbery, etc.). We conducted a principal components analysis; the first component explained 69% of the variance across the 13 variables and was dichotomized at the median to define a binary variable for higher vs. lower crime. Census tract crime variables were merged to each patient’s census tract of residence at diagnosis. We examined associations with molecular subtypes based on the 2017 European Leukemia Network prognostic score categories, receipt of high-intensity induction, receipt of stem cell transplantation, and survival.

Results: NHB patients were significantly more likely than NHW patients to reside in a high-crime neighborhood (87% vs. 13%, p<0.0001). Based on a tract-level composite measure of disadvantage, higher crime neighborhoods tended to be more socioeconomically disadvantaged (p<0.0001). Patients in higher-crime neighborhoods were less likely to be privately insured, but no other differences in patient characteristics according to age, gender, body mass index, or comorbidities were apparent. Patients residing in higher crime census tracts were more likely to have a high-risk karyotype (47% vs. 33%, P=0.03) and more likely to have a pathogenic TP53 mutation amongst those that had testing done (16% vs. 5%, p=0.10). Secondary AML based solely on chromosomal changes was more frequent for patients residing in higher vs. lower crime neighborhoods (30% vs. 18%, p=0.03). Patients residing in higher crime neighborhoods qualitatively had lower rates of high-intensity induction therapy (65% vs. 70%) and lower rates of allogeneic stem cell transplant (29% vs. 41%, P=0.08). These associations were attenuated with adjustment for age, race/ethnicity, and tract sociodemographic characteristics. There was no association with survival based on the current limited sample size. Data collection is ongoing, and our results will be updated as the sample size increases.

Conclusion: Our ongoing work suggests a link between neighborhood-level crime and AML biology and outcomes. Previously published mechanisms that may underlie this link between violence and leukemia include systemic inflammation and monocyte activity in response to neighborhood violence, as well as alterations in cortisol metabolism and glucocorticoid receptor binding in lung cancer tumors in response to violence, leading to more aggressive biology (Miller et al., 2022) (Heath et al., 2024).

Understanding the individual response to neighborhood exposures and its effect on leukemia development and outcomes is novel and may contribute to the racial disparity in leukemia survival.