NEW YORK – Race-based survival disparities in patients diagnosed with acute myeloid leukemia have widened over time, even as survival rates overall have trended toward improvement. This is one of the troubling findings of a study presented Sunday during the American Society of Hematology's annual meeting.
The study, led by Bhavana Bhatnagar of the Ohio State University and simultaneously published Friday in the journal Cancer Discovery, evaluated a variety of factors, including distinct genetic features among Black AML patients that may be contributing to these disparities.
Bhatnagar and colleagues took a two-pronged approach to unpacking the differential outcomes among Black and white patients with AML. In the first part of the study, they used population-wide data from the National Cancer Institute's Surveillance Epidemiology and End Results (SEER) Program to home in on the effects of age, sex, income level, healthcare coverage, and other socioeconomic and demographic factors on AML survival. In the second part of the study, the researchers evaluated the cytogenic and molecular features of Black and white patients treated according to protocols established by Cancer and Leukemia Group B or Alliance for Clinical Trials in Oncology research cooperatives, with the goal of determining whether these features, if distinct, influenced patient outcomes.
From the SEER-based portion of the study, which specifically included a cohort of 1,356 Black patients and 8,074 white patients with AML below the age of 60, the researchers found that 34 percent of Black patients lived for three years after their AML diagnoses compared with 43 percent of white patients. Black patients had a 27 percent higher risk of death than did white patients even after the researchers adjusted for the influence of factors such as age, sex, poverty levels, whether patients lived in a metropolitan area, and the decade during which they were diagnosed. In patients over the age of 60, the survival disparities also persisted according to race, but to a lesser degree than in the younger patients.
In the molecular analysis portion of the study, the researchers found that among patients younger than 60, certain cancer-linked gene mutations showed up at different rates in Black and white patients. Specifically, NPM1 mutations occurred in 25 percent of Black patients versus 38 percent of white patients; WT1 mutations occurred in 3 percent of Black patients versus 10 percent of white patients; IDH2 mutations occurred in 17 percent and 8 percent of Black and white patients, respectively; and PIK3CD mutations occurred in 4 percent and 1 percent of Black and white patients, respectively.
Notably, in the portion of the study that evaluated patients treated according to the Cancer and Leukemia Group and Alliance research protocols, patients' initial remission rates — 71 percent — were consistent between both Black and white patients. This, the researchers noted, reflects in part the fact that, in these controlled clinical trial settings, all patients were treated under the same protocols, receiving anthracycline-based induction therapy and no allogeneic stem-cell transplants. The patients' responses to induction therapy were similar regardless of their race. Within the first 30 days following therapy, rates of death were also comparable.
Upon longer-term follow up, however, these similarities dissolved. Three years after diagnosis, for instance, the disease-free survival rate among Black patients was 25 percent versus 38 percent among white patients, and the survival rate was 29 percent versus 42 percent, respectively. Seventy-one percent of Black patients experienced disease relapse, moreover, versus 59 percent of white patients.
The fact that within these controlled study protocols patients' outcomes started off as similar between Black and white patients but diverged significantly down the road stood out to researchers because it suggested that there were factors yet unaccounted for that were contributing to these disparities.Â
"We were very surprised to find that clinical trial enrollment did not alleviate survival disparities for younger Black AML patients," Bhatnagar reflected during a press conference on Friday. "That [difference in longer-term outcomes] lets us know that something happens between the time that these patients achieve remission and the time that they pass away that is unaccounted for, [and] trying to understand that gap will be critical in order to address this disparity, because I believe it to be multi-factorial."
Ultimately, Bhatnagar and co-authors wrote in the Cancer Discovery paper that "Black race was associated with worse overall survival compared with white race, with Black patients having a 40 percent higher likelihood of death compared with white patients after adjustment for white blood cell count, age, internal tandem duplication (ITD) of the FLT3 gene, and NPM1 mutation status." Black race, in other words, was an independent prognosticator of poor survival outcomes.
Examining further the impact of patients' race in the context of molecular features, the researchers noted that survival disparities between Black and white patients were particularly striking among those patients who harbored NPM1 mutations. Usually, NPM1 mutations without concurrent FLT3-ITD with high allelic ratio is a biomarker for improved prognosis in AML. However, this was significantly less true for the Black patients assessed in the study than it was for the white patients.
After three years, the disease-free survival rates and overall survival rates for white patients with NPM1 mutations were 41 percent and 47 percent, respectively, whereas these two rates were 23 percent and 17 percent for Black patients. In contrast, there were no significant differences observed between Black patients and white patients who did not have NPM1 mutations and who were considered to have a favorable disease risk score as determined by the European LeukemiaNet classification.
Based on these findings, the researchers wrote, "the poorer outcomes of Black AML patients may be driven, at least in part, by the poor survival of the NPM1-mutated, FLT3-ITD low or no allelic frequency patients."
The researchers noted that their findings should be further validated in larger patient cohorts and that follow-up studies should factor in disease characteristics beyond those that Bhatnagar and colleagues analyzed.
"Our analyses of both the SEER registry and the Alliance data were all retrospective and as a result of that are imperfect," Bhatnagar said. "Future controlled, prospective studies that are designed to look at these biological differences between younger Black AML patients and younger white AML patients are definitely needed in order to support our findings."
Bhatnagar expressed hope that future research efforts might consider assessing differences in gene mutations that aren't commonly studied in AML, because these could potentially be mutated at different frequencies between Black and white AML patients. She also urged researchers to conduct multidisciplinary studies to address the effects of socioeconomic disparities and structural racism alongside genetic differences.
"There are so many touchpoints that our patients have during the course of their treatment, and they encounter different forms of bias, which I think ultimately can contribute to their poor outcomes," she said. "Structural racism is very hard to study in the setting of clinical trials, and we would need very skilled social scientists to be able to help us design a study that's better able to assess it. We definitely believe that there is some component of [structural racism,] which is likely contributing to poorer survival outcomes in younger Black AML patients."
Characterizing the factors that contribute to race-based cancer disparities, both on a societal level and a molecular level, is not an easy task. Bhatnagar and colleagues' study underscores not only how important it is to approach this topic from both angles, but also how crucial it is for cancer clinical trials at large to reflect a diverse population of patients — patients who all deserve to benefit from treatment advances in an equitable way.
Worth noting in this study — or any study that examines race in the context of health disparities — is the fact that race, in and of itself, is a social construct that does not carry with it any intrinsic biological differences. The different disease features noted in this study correlated with patients' race, but were not features of race.
"We can't lose focus of the fact that race is a social construct," said Chancellor Donald of Tulane University in a discussion of the study findings. "But something else that often isn't said as we evolve as a society … is that we don't stress enough that race, although a social construct without any biologic basis, is permanent. This is something that will affect that person's life, and no matter what, the race that they are perceived to be or they are designated to be, is going to shape how they experience different outcomes, as well as difficulties they may face in their lives."