Better Therapy Selection for Childhood Leukemia
Despite decades of optimization of treatment protocols, the prognosis for acute myeloid leukemia in children (pediatric AML, pedAML) remains poor for many patients. A research team from St. Anna Children’s Cancer Research Institute (St. Anna CCRI), the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, the Medical University of Vienna, and St. Anna Children’s Hospital has now succeeded in developing a method for the early detection of resistance mechanisms in pediatric AML, using cutting-edge imaging, molecular methods, and computer-assisted data analysis,. The study has been published in the journal Cell Reports Medicine.
Pediatric acute myeloid leukemia (pediatric AML) is one of the most aggressive cancers in children. It arises when immature precursor cells in the bone marrow undergo genetic changes that block their normal maturation into functional blood cells. Instead, defective cells multiply uncontrollably, displacing healthy blood production and causing severe symptoms such as anemia, increased susceptibility to infections, bleeding tendency, and organ failure.
Unlike acute lymphoblastic leukemia, which occurs more frequently in children, pediatric AML is biologically more heterogeneous and in part more difficult to treat. Although survival rates have improved through advances in chemotherapy and stem cell transplantation, the prognosis for many patients remains unsatisfactory: some do not respond to standard therapies or suffer relapses. A new study published in Cell Reports Medicine now shows that functional imaging and molecular characterization can be combined into a tool that detects therapy resistance already at diagnosis.
This study is the result of particularly close collaboration among the research teams of Kaan Boztug, Michael Dworzak, and Giulio Superti-Furga—a joint effort between basic research and clinical practice that received €585,000 in funding from the Austrian Science Fund (FWF) under the Clinical Research program for the project Linking ex-vivo chemosensitivity, treatment and pathway activations for a deeper understanding of pediatric AML (ExTrAct-AML).
First author Ben Haladik, a PhD student in Kaan Boztug’s research group, together with the team further developed a platform for testing drug activity. It is based on the Pharmacoscopy method developed at CeMM for high-resolution imaging, combined with artificial intelligence and comprehensive molecular analysis. Using 45 patient samples, they demonstrated that robust predictions of therapy response and relapse risk can be derived.
Molecular Profile as a Key to Prognosis
Leukemia cells from blood or bone marrow samples are treated in the laboratory with various drugs, and then observed under the microscope to see whether they die or remain resistant. This is done on a large scale and fully automatically: using deep-learning algorithms, the effect of each drug on hundreds of thousands of cells is analyzed in parallel. Combined with genetic and epigenetic data, this yields a detailed “chemosensitivity profile.”
Vulnerable to Known Drugs
Clear differences emerged between risk groups and even among subpopulations of cells that escape standard therapy. Particularly striking was a stem-cell-like form of leukemia that proved insensitive to conventional chemotherapy but was vulnerable to new combinations of known agents such as BCL2 and MDM2 inhibitors or HDAC blockers. The results show that functional analyses of this kind can further improve therapy prognosis for pediatric AML. While mutations provide important clues, the real clinical relevance lies in how leukemia cells respond to drugs.
This is precisely where the new method comes in: it makes the functional level visible and allows a direct link between molecular profile and actual therapy response.
From Research to the Clinic
This form of functional precision medicine has the potential to fundamentally change the treatment of pediatric AML. It complements genetic diagnostics and the detection of minimal residual disease—currently the most important tools for risk assessment—by adding a dimension that directly depicts drug response. This brings the possibility closer of identifying high-risk patients already at diagnosis and offering them targeted new therapies.
First author Ben Haladik explains the methodology: “We have created a connection between molecular biological analyses, bioinformatics methods, and artificial intelligence that should provide a basis for further research toward better treatment options.”
Senior and corresponding author Kaan Boztug sees the study as a societal mission for the future: “Our study is the first to show that such ex-vivo drug tests can help us identify, at an early stage, patients whose leukemia cells are particularly resistant to standard therapy. Especially for these patients, we can then use the method to find targeted therapy options for pediatric AML. With our study we also position ourselves in a previously little-noticed field—the application of AI in childhood cancer research—as a significant player in European pediatric cancer research.”
A Milestone
“At CeMM we developed Pharmacoscopy, an imaging-based approach for functional single-cell precision medicine—a technology that enables true personalized medicine in cancer treatment. In the present study, this technology was further developed and successfully tested for the first time in the clinical diagnosis of pediatric AML. This is an important milestone toward implementing such methods on a larger scale for the benefit of pediatric patients,” says Giulio Superti-Furga, co-senior author of the study.
“The results of our study open up a completely new approach to treating pediatric AML. By detecting resistance at the time of diagnosis, we lay the foundation for using therapies in a much more targeted and individualized way in the future. This means we can identify high-risk patients early and offer them more precise treatment strategies—an essential step toward sustainably better chances of cure,” adds Michael Dworzak, head of the “Immunodiagnostics” research group at St. Anna CCRI and deputy medical director at St. Anna Children’s Hospital. The results presented are based on a retrospective cohort. The next step will be prospective clinical studies in which the method will be applied in real time and compared with actual disease progression.