Advancing Treatment Options for AML: Targeting Leukemia Stem Cells
Acute Myeloid Leukemia (AML) remains one of the most challenging blood cancers to treat, with a high rate of relapse due to therapy-resistant leukemia stem cells (LSCs). This project aimed to identify new treatment strategies that specifically target LSCs, improving patient outcomes.
One promising approach involves using the body’s own immune system—particularly Natural Killer (NK) cells, a type of immune cell capable of attacking cancer cells while sparing healthy ones. However, previous attempts to treat AML by injecting NK cells into patients have had limited success. Research from our lab uncovered a key reason: not all AML cells are equally vulnerable to NK cells. Only those that express specific molecular markers—NKG2D-ligands, which activate NK cells—are highly susceptible to attack. In contrast, AML cells lacking these ligands, including most leukemia stem cells, were highly resistant.
A New Understanding of Leukemia Stem Cells
We discovered that the expression of NKG2D-ligands, and therefore the susceptibility of AML cells to NK cells, is linked to how mature (or differentiated) the leukemia cells are. LSCs often remain in an immature, stem-like state, which makes them highly resistant to immune attack. These resistant cells, which closely resemble normal hematopoietic stem cells, were classified as canonical LSCs.
However, we identified a previously overlooked subgroup of AML patients whose LSCs naturally expressed high levels of NKG2D-ligands, making them much more vulnerable to NK cells. These cells, which we termed non-canonical LSCs (NcLSCs), resemble a more mature stage of blood cell development called the myelomonocytic precursor stage. Importantly, this subset of patients could be easily identified using a simple and rapid flow cytometry assay.
Translating Discoveries into Treatment
Animal studies showed that patients with NcLSCs could benefit significantly from NK cell therapy without additional interventions to induce NKG2D-ligand expression. We also discovered that, beyond NKG2D-ligands, the interaction between ICAM-1 on NK cells and LFA-1 on AML cells played a crucial role in how NK cells recognize and attack these leukemia stem cells. Blocking this interaction significantly reduced NK cell effectiveness, highlighting a key mechanism of immune evasion in AML.
Despite its potential, NK cell therapy remains expensive, is not yet a standard treatment, and lacks the necessary infrastructure in Germany. Fortunately, our research revealed that NcLSCs are also highly sensitive to existing, widely available cancer drugs, including MEK1/2 inhibitors (commonly used in melanoma) and Arsenic Trioxide (ATO) (used in acute promyelocytic leukemia) when combined with standard chemotherapy. This breakthrough has led to a patent application for a new treatment strategy, bringing us one step closer to clinical application.
Predicting Treatment Response with the MAC-Score
Another major challenge in AML treatment is predicting which patients will respond to existing therapies. One such therapy, Venetoclax (VEN) combined with hypomethylating agents (HMA), is commonly used for older or frail AML patients. However, around 30% of patients do not respond, and there has been no reliable way to predict outcomes before starting treatment.
Our research found that non-responding patients had LSCs in a unique anti-apoptotic state, meaning they did not rely on BCL-2 (the target of Venetoclax) for survival. To translate this discovery into a clinical tool, we developed a flow cytometry assay that measures three key anti-apoptotic proteins: BCL-2, BCL-xL, and MCL-1. The ratio of these proteins, which we named the Mediators of Apoptosis Combinatorial-Score (MAC-Score), was able to predict Venetoclax response with over 97% accuracy in a retrospective study of more than 110 AML patient samples.
Clinical Impact and Future Directions
The MAC-Score outperformed genetic markers of Venetoclax response and provides a fast, affordable, and easy-to-implement test, delivering results within 3-4 hours. The method has been patented, validated at the University Hospital Heidelberg, and is currently being expanded to other clinical centers. Future efforts aim to commercialize and adapt MAC-Score technologies for other blood cancers, potentially transforming AML management and improving patient outcomes.
Our findings have been published in multiple peer-reviewed journals, presented at international conferences, and highlighted in various scientific outlets. By bridging the gap between laboratory discoveries and clinical application, this work has the potential to revolutionize personalized treatment for AML patients worldwide.
