21/10/2026

Print Page

A New Platform of Robust Patient-Derived AML Models for Translational Research

Acute myeloid leukemia (AML) is an aggressive blood cancer with a dismal prognosis. Therefore, researchers aim to identify new treatment options. However, as AML is a very heterogenic cancer, it is difficult to study this disease. Conventional leukemia cell lines are easy to grow, but often no longer reflect the biology of patients’ tumors, and many subtypes are not reflected at all by cell line models. In this study, researchers from the lab of Prof. Irmela Jeremias, LMU Munich, created serially transplantable and robust patient-derived xenograft (STAR-PDX) models. These were generated by transplanting patients’ leukemia cells into immunodeficient mice, where they grow in a physiological bone-marrow environment.

*PDX Models of Acute Myeloid Leukemia for Translational Cancer Research © AG Jeremias

Starting with more than 100 patient samples, originating from several DKTK partner sites, the team selected models with special criteria to allow their use for complex molecular studies and state-of-the-art preclinical treatment trials. These criteria involved serial re-transplantation in mice, survival of freezing and thawing cycles, reliable growth, and capacity for genetic modification. The resulting *PDX collection covers different age groups and AML subtypes, including clinically relevant subtypes for which no established cell lines exist, such as AML with IDH1/2 mutations or combined NPM1 mutation and FLT3-ITD. Detailed analyses showed that the models preserve patients’ genetic changes, gene activity, DNA methylation patterns, and cell-surface markers better than commonly used AML cell lines.


By introducing a special transgene into the cells, the scientists can track leukemia growth and treatment response in real-time. Different models responded very differently to chemotherapy, closely resembling the varied responses seen in patients. In an exceptionally long therapy trial lasting about one year, repeated chemotherapy did not cause obvious resistance, but surviving leukemia cells grew progressively more slowly. This unexpected result suggests that treatment may alter leukemia growth, a finding which cannot be captured in short-term experiments.


This *PDX model collection is a unique and extremely valuable tool for leukemia research, as it provides renewable, standardized models for testing therapies across genetically diverse AML cases preceding their use in humans in clinical trials. It may help to identify which patient groups are most likely to benefit from a new drug and make preclinical results more reliable. 


The platform brings together more than a decade of pioneering work by the Jeremias lab and opens new avenues for translational leukemia research. It combines the lab’s expertise in genetic engineering, leukemia biology, and sophisticated in vivo preclinical trials monitored by real-time in vivo imaging, in one shared resource. The models have already supported more than 48 publications, many of them in collaboration with DKTK partners, and are available through CancerModels.org. Wider use should strengthen collaborations and reproducibility to drive the translation of biological discoveries into better-targeted AML therapies.

 

Original Publication

Further information on PDX Models