African American Race Associated With Poor Survival in Younger Patients with AML

December 7, 2020
Hayley Virgil

The most important factor associated with poor survival among patients below 60 years of age with acute myeloid leukemia was self-reported black race in a SEER analysis.

The most important factor associated with poor survival among patients below 60 years of age with acute myeloid leukemia (AML) was self-reported black race in a Surveillance Epidemiology and End Results (SEER) analysis, according to the findings presented during the 2020 ASH Virtual Annual Meeting.

Although overall survival (OS) has improved over the past 30 years for patients with AML, survival disparities between black and white patients has widened over time (P <.001). A non-statistically significant difference in survival for patients diagnosed with AML between 1986-1995 (white patients, n = 1365; black patients, n = 160; P = 0.19), although this achieved significance between 1996-2005 (white patients, n = 2994; black patients, n = 480; P = .004). The gap has become even more pronounced in the most recent decade (white patients, n = 3715; black patients, n = 716; P <.001).

Younger black patients (n = 1356) with AML were also found to have worse survival outcomes compared to their white counterparts (n = 8074), with 3-year survival rates of 34% for younger black patients versus 43% in younger white patients (P <.001). Consistent with previous studies, black patients with AML were found to be younger at the time of diagnosis compared with white patients (P <.001). Additionally, a higher number of black patients were found to reside in metropolitan areas (P <.001) and to be either uninsured or on Medicaid (P <.001) versus white patients. A higher percentage of black patients were found to live below the poverty line versus white patients (P<.001).

“Multivariable analyses for OS in younger patients with AML showed that black patients were found to have a 27% higher risk of death compared with white patients after adjustment for age, sex, metropolitan area of residence, measure of poverty, and decade of diagnosis,” Bhavana Bhatnagar, DO, an osteopathic physician and an associate professor-clinical in the Leukemia Research Program at The Ohio State University Comprehensive Cancer Center—The James, said during a presentation.

The study had 3 main objectives: to determine whether racial disparities in clinical outcomes continue to persist between black and white patients with AML, to evaluate whether differences in the cytogenetic and mutational landscapes exist between these populations when receiving similar treatments for their disease, and to understand whether genetic factors contribute towards the survival outcomes of black patients with AML.

Two datasets were analyzed by investigators in order to answer these questions, the first being a SEER registry, which included data from 25,523 non-Hispanic black and non-Hispanic white patients aged 18 years or older who had been diagnosed with AML between 1986 and 2015. Data on patients with acute promyelocytic leukemia were excluded from the analysis. Investigators examined demographic, socioeconomic, and survival differences between the 2 patient subsets.

The second analysis, from the Cancer and Leukemia Group B/Alliance, focused on a subset of 1339 black and white patients with AML, based on self-reported race, who received intensive treatment with cytarabine-/anthracycline-based induction chemotherapy on Alliance protocols between 1986 and 2015. These patients also underwent next-generation sequencing studies of 81 genes to evaluate differences in mutations and survival. Notably, neither of the patient subsets received allogeneic stem cell transplant in first remission, as mandated by the treatment protocols.

Results from the second analysis of younger black (n = 72) and white patients (n = 777) with AML who were treated on Alliance protocols indicated no significant differences in early death rates (10% versus 46%, respectively; Pa = .02) and complete remission rates (71% vs 71%; Pa= 1.00). “This indicates similar initial response to induction treatment and likely comparable performance status and early complication rates,” said Bhatnagar. Relapse rates were slightly higher in black patients vs white patients, at 71% versus 59%, respectively, but this difference did not reach statistical significance (Pa= .14).

However, black patients experienced significantly worse disease-free survival (DFS; P = .02) and OS (P = .02) compared with their white counterparts; this may indicate that other factors beyond treatment selection are impacting the survival disparity between these patient subsets, Bhatnagar explained.

From there, investigators looked for underlying molecular differences between the 2 subsets to determine whether differences in underlying disease biology may contribute to survival disparities. A targeted next-generation sequencing analysis of 81 leukemia-associated genes indicated that NPM1 (27%), FLT3-ITD (24%), and DNMT3A (24%) were the most commonly mutated genes detected in black patients with AML. Other commonly mutated genes included IDH2 (18%), NRAS2 (16%), TET2 (15%), IDH1 (12%), and TP53 (11%).

When looking at the mutational profiles of black and white patients with AML, investigators found differences in mutational frequency between the populations. NPM1 (P = .04) and WT1 (P = .05) were found to be more common in white patients (38% vs 10%, respectively) compared to black patients (25% vs 3%, respectively). IDH2 was found to more commonly be mutated in black patients (17%) compared with white patients (8 %; P = .03).

“Notably, no differences were seen between the assignment of favorable, intermediate, or adverse genetic risk groups based on cytogenetic findings and selective gene mutations, as defined by the 2017 European LeukemiaNet Guidelines,” Bhatnagar noted.

In multivariable analyses of OS, investigators found that black race was associated with worse OS versus white race. Specifically, black patients had a 40% higher likelihood of death versus their white counterparts after adjustment for age, white blood cell counts, and FLT3-ITD and NPM1 mutation status. “Thus, our analysis suggests that black race represents an independent adverse prognostic factor in young patients with AML,” Bhatnager said.

Additionally, younger black patients with AML were found to have less mutations in the NPM1 gene, which is a common gene mutation in patients with this disease, according to Bhatnager. “In the absence of FLT3-ITD, this [mutation] associates with a favorable prognosis,” she said. In an analysis of patients who harbor the prognostically favorable NPM1 mutation, investigators observed that black patients had notably worse DFS (P = .009) and OS (P<.001) compared with white patients.

Because FLT3-ITD is known to co-occur with NPM1 mutations in some patients and offsets the favorable prognostic value of NPM1 mutations, investigators went on to examine the outcomes of young black patients with favorable-risk genotypes. Even in this patient population, it was observed that black patients with AML had worse DFS (P = .02) and OS (P <.001) versus white patients. Bhatnagar noted that the numbers reported were small and require further corroboration.

"While I don't think we can currently make any definitive conclusions, we have demonstrated in a large group of patients that there are biological differences within the actual leukemia in black and white patients,” Bhatnagar said.

In both univariable and multivariable outcome analyses to evaluate clinical or molecular features that impact clinical outcome, investigators found that FLT3-ITD (HR, 1.95; 95% CI, 1.06-3.57; P = .03) and IDH2 (HR, 2.17; 95% CI, 1.27-4.81; P = .008) were the only mutations that are associated with a higher risk of death in black patients. FLT3-ITD and IDH2 are now targetable with small molecule inhibitors, which is important, Bhatnagar said.

Bhatnagar told OncLive that the findings of this analysis need to be corroborated in larger prospective studies in the future, as both the SEER and Alliance analyses were retrospective.

“Future controlled prospective studies that are designed to look at these biological differences between young black and white patients with AML are needed in order to further support our findings,” concluded Bhatnager. “Efforts also need to look at other gene mutations that we don't commonly study in AML but could potentially be mutated at different frequencies between black and white patients with AML. Lastly, these studies will need to be multidisciplinary in nature to address some of the socioeconomic and structural racism questions that our study has hopefully raised.”


Bhatnagar B, Zhao Q, Fisher JL, et al. Poor treatment outcomes of young (<60 years) African American patients (Pts) diagnosed with acute myeloid leukemia (AML) (Alliance). Presented at: 2020 ASH Annual Meeting & Exposition; December 5-8, 2020; Virtual. Poster 6.