Racial and Ethnic Related Socioeconomic Disparities Influence Survival Outcomes in AML

Census tract socioeconomic status information demonstrated significant disparities between survival outcomes of non-Hispanic white, non-Hispanic black, and Hispanic patients with acute myeloid leukemia (AML) in the Chicago metropolitan area, according to findings from a multilevel analysis presented virtually during the 2020 ASH Annual Meeting.

Moreover, socioeconomic tract information, including disadvantage and affluence, accounted for 81% of the black-white disparity in AML-related death within this patient population (HR decreased from 1.58 to 1.11 with adjustment for tract socioeconomic status).

Treatment differences related to induction treatment and allogeneic stem cell transplant (allo-SCT) accounted for 41% of the disparity.

“Interestingly, comorbidities and molecular disease-specific differences did not mediate black-white or non–white-white disparities in leukemia-specific death,” said Irum Khan, MD, senior study author and an assistant professor of medicine at the University of Illinois College of Medicine, in a virtual presentation of the data.

Ultimately, these results represent a need to develop validated measures of structural violence within oncology.

The study represents the first of its kind to incorporate clinical and molecular data from patients with AML with neighborhood characteristics and treatment patterns. The goal of the study was to shed light on the effect of racial and ethnic disparities on survival outcomes of patients with AML.

Structural is defined as institutionalized social forces, and violence suggests the systematic advantage or disadvantage these social forces impose on some groups. As such, structural violence in the health care field implies that there are social processes, such as racism, social class, and heteronormativity, that produce and perpetuate health inequalities.

Although disparities exist throughout the health care system, decreased overall survival had been reported among non-Hispanic black and Hispanic adults and adolescent and young adult patients with AML. Notably, these disparities are somewhat counterintuitive as these patient populations have a higher prevalence of favorable cytogenetics and younger age at diagnosis compared with non-Hispanic white patients with AML.

From 1991 to 2008, survival has improved by 16.4% for non-Hispanic white patients with AML vs 10.8% for other patients. Contributing factors to the disparity include:

• Non-Hispanic black patients are less likely to receive intensive chemotherapy and allogeneic stem cell transplant as treatment for AML compared with non-Hispanic white patients
• Non-Hispanic black and Hispanic patients with AML are more likely to be treated at heath care facilities that see low volumes of patients with AML
• Non-Hispanic white patients with AML report higher rates of early mortality (death within 60 days of diagnosis)

At diagnosis, factors including age, gender, institution, and socioeconomic status play a role in AML-related morbidity and mortality. Specifically, socioeconomic status can influence clinical factors, such as body mass index and Charlson comorbidity index, molecular factors, such as European Leukemia Net (ELN) score, high-risk mutations, and secondary leukemia, and health care system factors, such as clinical trial enrollment and health insurance. These factors then contribute to treatment options, intensive care unit admission, and ultimately, AML-related death.

The multilevel analysis aimed to bridge a gap in literature regarding these discrepancies in AML and shed light on how structural violence, particularly neighborhood socioeconomic status, contributes to AML-specific survival.

The study evaluated data from adult non-Hispanic white, non-Hispanic black, Hispanic, or other adult patients with AML (non–acute promyelocytic leukemia) who were diagnosed between 2012 and 2018. Patient data was accrued from 6 academic cancer centers in Chicago: the University of Illinois Chicago (n = 126), Loyola University (n = 229), Northwestern University (n = 151), the University of Chicago (n = 150), Rush University (n = 100), and the John H. Stroger, Jr. Hospital of Cook County (n = 66).

“The reason we opted to do this in our city was because Chicago is one city with a high level of segregation,” said Khan, who highlighted that Chicago is the third most segregated city in the United States. “It did provide a unique opportunity to study this variable in some detail and how it interacted in a multilevel analysis with other leukemia-specific markers of prognosis.”

The Federal Financial Institutions Examination Council’s Geocoding/Mapping System was used to collect census tract data including concentrated disadvantage factors, such as families with incomes below the poverty line, families receiving public assistance, unemployed individuals, and female-headed households with children, as well as concentrated affluence factors, such as families with incomes greater than $75,000, adults with college education or more, and labor force in professional or managerial occupations. 

Overall, 497 patients were non-Hispanic white, 126 were non-Hispanic black, 117 were Hispanic, and 82 were considered other.

“The sample is deliberately representative of a diverse socioeconomic background and intentionally designed to represent the entire population [of patients with AML in Chicago].”

Patient characteristics varied between racial/ethnic background. Overall (n = 822), the majority of patients (n = 459) were above the age of 60 at diagnosis. However, more Hispanic patients (25%) were 18 to 39 years old at diagnosis compared with patients who were non-Hispanic white (9%), non-Hispanic black (15%), or other (11%).

The majority of patients had Charlson index scores of 2 or greater (n = 380). However, the frequency of Charlson index scores of 2 or greater was lower in Hispanic patients (48%) vs non-Hispanic white patients (65%), non-Hispanic black patients (69%), and other (70%). More Hispanic patients had Charlson index scores of 0 or 1 compared with the other patient subgroups.

Furthermore, morbid obesity was more prevalent in non-Hispanic black and Hispanic patients (22% and 19%, respectively) vs non-Hispanic white and other patients (11% and 14%, respectively). Also, regarding payer information, private insurance was held twice as commonly by non-Hispanic white patients compared with non-Hispanic black patients.

Regarding molecular characteristics (n = 810), favorable ELN risk scores (low or intermediate risk) were more prevalent among Hispanic patients (71%) vs non-Hispanic black patients (58%) and non-Hispanic white patients (62%). Secondary AML was reported in 31%, 51%, and 41% of patients, respectively. p53 mutations were identified in 9%, 26%, and 12% of patients, respectively; however, only 281 patients had data regarding p53 mutational status, so the difference was not deemed statistically significant.

Prior treatment patterns showed that Hispanic patients had the highest rate of intensive induction therapy (80%) compared with 68% in non-Hispanic black and non-Hispanic white patients. A disparity in intensive induction therapy complication rates was confirmed; Hispanic and non-Hispanic black patients had significantly higher rates of ICU admission during induction chemotherapy (41% and 38%, respectively) vs 26% of non-Hispanic white patients.

Additionally, 47% of non-Hispanic white patients, 22% of non-Hispanic black patients, and 42% of Hispanic patients underwent allo-SCT after induction chemotherapy.

Socioeconomic tract characteristics revealed differences in tract disadvantages and affluence between racial/ethnic groups in metropolitan Chicago. Forty-six percent of non-Hispanic white patients fell into the lowest third of tract disadvantage vs 7% of non-Hispanic black patients and 10% of Hispanic patients. In the middle third, these rates were 38%, 12%, and 33%, respectively. Notably, in the highest third of tract disadvantage, the reported percentages were 15%, 82%, and 57%, respectively (overall tract disadvantage, P < .0001).

Tract affluence revealed that 20% of non-Hispanic white, 54% of non-Hispanic black, and 72% of Hispanic patients fell into the lowest third group; 37%, 33%, and 17% fell into the middle third; and 43%, 13%, and 10% fell into the highest third, respectively (overall tract affluence, P < .0001).

The majority of non-Hispanic black patients (89%) lived in primarily African American census tracts (8% to 100%) vs 19% of non-Hispanic white patients and 38% of Hispanic patients. Conversely, 49% of non-Hispanic white patients, 5% of non-Hispanic black patients, and 8% of Hispanic patients resided in predominantly white census tracts (83% to 99%).

“I believe that concentrated disadvantage is common to a lot of urban centers and although yes, this is a Chicago-wide study, we think it is broadly applicable to most metropolitan centers,” explained Khan.

Future research efforts are needed to continue to disentangle the influence socioeconomic status has on cancer outcomes. For example, mixed-methods studies and community engagement is needed to shed light on social and economic barriers that exist within the cancer community. Additional studies are also needed to inform the biologic pathways that are relevant to structural violence, as well as to identify prognostic factors that can effectively measure structural violence, Khan concluded.

Reference

Abraham I, Rauscher G, Patel A, et al. The role of structural violence in acute myeloid leukemia outcomes. Presented at: 2020 ASH Annual Meeting; December 5-8, 2020; virtual. Abstract 217.