Even when patients with acute myeloid leukemia (AML) appear to be in remission, small numbers of cancer cells can remain in the body. These lingering cells—known as measurable residual disease (MRD)—are one of the strongest predictors of whether the leukemia will return. To detect them, clinicians often use multiparameter flow cytometry, a technique that analyzes proteins on the surface of individual cells to distinguish abnormal leukemia cells from normal blood cells. This approach can identify extremely rare leukemic cells that are indistinguishable by microscopy.
However, MRD testing is not always straightforward. Leukemia cells can vary widely in the types of surface markers they display and in how closely they resemble normal stem or progenitor cells. These differences create distinct “immunophenotypes,” or marker patterns, that help define abnormal cells. Importantly, some MRD-positive patients relapse quickly, while others have outcomes similar to patients with no detectable MRD. This variability suggests that simply labeling MRD as present or absent may overlook important biological differences that influence patient prognosis.
“Results from MRD testing are increasingly used in the care of patients with AML—not only for prognostication but also to guide treatment decisions—even though it is well recognized that MRD testing is far from perfect,” says Dr. Roland Walter of the Translational Science and Therapeutics Division. In a recent study published in HemaSphere, Walter and colleagues set out to determine whether more detailed analysis of leukemia cell immunophenotypes could improve the prognostic value of MRD testing.
Their findings suggest that two key factors matter: the type of leukemic cells present and the number of immunophenotypic abnormalities they carry. Patients whose residual leukemia cells displayed a greater number of abnormalities had significantly worse outcomes. In contrast, the researchers identified a subset of MRD-positive patients whose leukemia cells showed relatively few abnormalities—and whose outcomes were strikingly similar to patients with no detectable MRD. “Essentially, our study allows us to reclassify some patients who test MRD-positive as having a better prognosis than would be predicted using traditional flow cytometric MRD assays,” Walter explains.
In this retrospective study, the team analyzed 1,215 adults with AML or myelodysplastic syndrome (MDS), a related group of bone marrow disorders that can progress to AML, who underwent allogeneic hematopoietic cell transplantation while in morphologic remission. Using multiparameter flow cytometry on bone marrow samples collected during the pre-transplant workup, clinicians assessed measurable residual disease (MRD) by identifying rare leukemia cells with abnormal combinations or levels of cell surface markers compared to normal blood-forming cells. In this study, any detectable population of such abnormal cells was considered MRD-positive, allowing detection of leukemia at levels far below what can be seen by standard microscopy. Using this approach, MRD was identified in 233 patients (19%).
The researchers next asked whether characteristics of the residual leukemia cells themselves could further refine risk prediction. Among MRD-positive patients, leukemic blasts fell into three groups based on their immunophenotype: non–stem cell-like, stem cell-like, or a mixture of both. Patients with mixed populations had the highest relapse risk, while those with stem cell-like blasts alone showed a trend toward somewhat better outcomes, suggesting that the biological identity of the residual leukemia cells may influence disease behavior.
However, the most striking differences emerged when the investigators quantified the number and severity of immunophenotypic abnormalities in MRD cells. By assigning each patient an overall MRD immunophenotype score, they found that outcomes separated dramatically at a threshold score of 4.5. Patients with higher scores—indicating leukemia cells with many abnormal features—had a 74% relapse risk at three years and very poor relapse-free survival. In contrast, patients with lower scores had outcomes remarkably similar to those with no detectable MRD. After adjusting for other clinical factors, a high immunophenotype score remained strongly associated with relapse risk, relapse-free survival, and overall survival.
Importantly, this pattern held across multiple analyses, including when using the commonly applied 0.1% threshold to define MRD positivity. Together, these findings suggest that not all MRD carries the same prognostic weight. Instead, incorporating the degree of immunophenotypic abnormality may help distinguish high-risk MRD from cases where the residual leukemia cells appear biologically less aggressive.
While MRD testing by flow cytometry is a powerful prognostic tool in AML, interpreting these data can be complex and requires significant expertise. “Approaches to MRD testing are not well standardized, and data analysis can be challenging,” notes Dr. Sindhu Cherian, Associate Director of the Hematopathology Laboratory at the University of Washington. Although this testing has been performed at the University of Washington for more than two decades—where it plays an important role in guiding clinical decision-making—relatively few centers currently offer it.
By focusing on the number of immunophenotypic abnormalities present in residual leukemia cells, the researchers’ scoring approach could simplify MRD interpretation and make results easier to standardize across institutions. “As data analysis evolves, this strategy could lend itself to artificial intelligence–assisted approaches,” says Cherian, potentially expanding access to MRD testing and enabling more consistent analysis across laboratories.
Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium Members Drs. Xueyan Chen, Megan Othus, and Roland Walter contributed to this research.
The spotlighted research was funded by the National Institutes of Health and the José Carreras/E. Donnall Thomas Endowed Chair for Cancer Research.
Le CM, Chen X, Kodaira S, Othus M, Gang M, Davis C, Basom RS, Cherian S, Walter RB. 2026. Immunophenotypic abnormality quantification refines multiparameter flow cytometry-based measurable residual disease testing in adults allografted for acute myeloid leukemia in morphologic remission. Hemasphere. DOI: 10.1002/hem3.70310.
Science Spotlight writer Jenny Waters is a postdoctoral research fellow in the Hsieh lab at Fred Hutch. She studies how mRNA translation coordinates bladder cancer transformation and metastasis by post-transcriptionally regulating expression of oncogenic proteins. Outside of the lab, Jenny enjoys spending time with her dogs, convincing her husband to join her on trail runs, and pretending every steep hill is just a "gentle incline."
For the Media