Targeting survival pathways in acute myeloid leukaemia

Acute myeloid leukaemia (AML) is the most common acute leukaemia in adults. Currently, less than half of the patients afflicted can expect a long-term cure. The poor cure rate reflects both the toxicity of the treatment (chemotherapy, stem cell transplantation), as well as the inherent ability of AML blasts to survive cytotoxic insults, eventually causing disease relapse. Apoptosis (programmed cell death) is a tightly-regulated process that plays many critical roles in normal physiological functioning. This process is subverted in many cancers, including AML, often contributing to the cancer’s ability to resist the effects of chemotherapy. Targeting specific proteins involved in the apoptotic process, in particular the BCL2 protein family members, represents a novel therapeutic approach, and has proven to be successful in malignancies such as chronic lymphocytic leukaemia (CLL). However, AML cells often express multiple pro-survival BCL2 proteins, which provide a redundancy of survival signals. This body of work demonstrates the dependence of AML cells on prosurvival proteins BCL2 and MCL1 in mediating cell survival. We also show that inactivating single pro-survival BCL2 proteins is unlikely to be a successful therapeutic strategy, and that rational drug combinations targeting non-overlapping survival pathways offer superior outcomes in vitro and in animal models. In particular, the combination of venetoclax (BCL2 inactivation) and standard chemotherapy agents is often synergistic in vitro. Conversely, singly targeting the pro-survival proteins BCL2 and MCL1 individually may lead to the acquisition of resistance mediated by compensatory alterations in the apoptotic pathway. In addition, the resistance to a specific BH3 mimetic may confer crossresistance to other BH3 mimetics targeting different pro-survival BCL2 proteins, and blunts the synergistic effects of venetoclax/chemotherapy combinations. Using cell line models, we find that systems lacking the tumour suppressor p53, as well as pro-death proteins Noxa/PUMA do not benefit from venetoclax/chemotherapy combinations. As p53 mutations are more common in relapsed AML, our work suggests that venetoclax/chemotherapy combinations are best utilized as upfront, rather than salvage therapy. Finally, we also demonstrate that simultaneous inactivation of a combination of BCL2 prosurvival proteins may result in complete eradication of AML in a mouse model. Specifically, concurrent BCL2 and MCL1 inactivation prolonged survival in mice xenografted with human AML, and in some cases, resulted in a cure. This represents an exciting development, indicating that a chemotherapy-free approach to curing AML is feasible. In summary, targeting apoptotic pathways in AML represents a promising treatment paradigm that merits further investigation. The findings of this body of work will hopefully pave the way for future clinical studies.