Fine-tuning prediction of response and survival in acute myeloid leukemia

Overall survival in AML patients remains poor. Selection of the right treatment for the right patient will contribute to better response rates and survival. Biological data associated with response and survival may inform development of novel treatment strategies. In this thesis, we generated data to optimize prediction of survival and treatment response in AML patients. We reviewed commonly used culture techniques for AML samples, that are at the basis of targeted therapy development and response-prediction experiments. We identified several pitfalls that should be taken into account when planning and interpreting data from ex vivo drug response experiments. With the described recommendations, we aim to maximize the reproducibility and meaningfulness of ex vivo experiments with primary patient AML cells. Although novel treatment strategies are of paramount importance to improve clinical outcomes in AML patients and the development of novel agents has surged, chemotherapy is likely to remain the basis of curative treatment for the coming years. To better understand response mechanisms towards currently used chemotherapeutic regimens, we investigated molecular processes that determine their cytotoxic effects. Our data suggest that distinct gene expression profiles are associated with ex vivo drug response, and may confer a priori drug resistance in leukemic cells. Gene function and expression may be affected by genetic mutations, or may be altered by other (epigenetic) mechanisms, independent of mutations. These genetic aberrations often have prognostic implications for AML patients. The frequency of TP53 mutations and their prognostic impact in pediatric AML were unclear, as well as the impact of differential expression of p53 pathway genes. Therefore, we analyzed TP53 alterations in bone marrow samples of 229 children with de novo AML. Our results indicate that TP53 alterations in children are rare, but that alterations in TP53 – and differential expression of its pathway genes – are associated with poor survival in pediatric AML. Mutations that frequently occur in AML, are internal tandem duplications of FLT3 (FLT3-ITD). Although an established adverse prognostic marker in adult AML, its precise role in pediatric AML was unclear, as well as the preferred measurement level (DNA or RNA/cDNA). Our results show that in pediatric AML, a high FLT3-ITD allelic ratio (FLT3-ITD-AR) is predictive for poor overall survival. Importantly, only RNA/cDNA-based FLT3-ITD-AR measurements – and not DNA measurements – have prognostic value in pediatric AML patients. Also in adults, we showed that the use of RNA/cDNA enables better detection of minor FLT3-ITD clones than DNA. This primarily impacts treatment selection and not prognostication as the additional identified FLT3-ITD clones only represent minor clones with low allelic ratios. However, more AML patients would be eligible for treatment with FLT3-TKIs using a cDNA-based approach. Next, profiled primary AML samples at the phosphoproteomic level to characterize FLT3-ITD AML and identify phosphorylation profiles associated with FLT3-TKI response. We demonstrated that ERK1/2 activity is associated with impaired response towards gilteritinib, and that modulation of the RAS-RAF-MEK-ERK axis may be a promising approach to improve responses and reduce resistance towards FLT3-TKIs. In future research, these analyses should ideally be performed directly on clinical specimens from patients treated with midostaurin and gilteritinib to provide more direct mechanisms of clinical response towards FLT3-TKIs. Curative pharmacological options remain limited for AML and come at a high cost of severe side effects, and thus we are still awaiting novel treatments that improve duration and quality of life of all AML patients. We anticipate that in the future, refined patient selection based on integral prediction models for response and survival, combined with novel targeted agents and/or combination strategies will maximize clinical benefit of currently available therapies, as well as of novel therapies in development.