Background Acute myeloid leukaemia (AML) is an aggressive blood cancer. In approximately 30% of the cases, driver mutations in the FLT3 gene are identified. FLT3 inhibitors are used in treatment of such patients together with cytotoxic drugs or (in refractory AML) as single agents. Unfortunately, resistance to FLT3 inhibitors limits their efficacy. Resistance is often due to secondary mutations in the gene encoding the molecular target. The gatekeeper mutation F691L confers resistance to specific FLT3 inhibitors such as quizartinib, but pexidartinib is much less resistance to this mutation. Pexidartinib alone is however sensitive to many other resistance mutations. In chronic myeloid leukaemia (CML), it has been suggested that rotation between drugs with a different landscape of resistance mutations might postpone the emergence of resistance. Methods We studied the effect of quizartinib and pexidartinib in AML cell lines that express FLT3 (MOLM-14 and MV4-11). Using a rotation protocol, we further examined whether the emergence of resistance could be postponed. Computational modelling was used to analyse the onset of resistance and suggest which mutations are most likely to occur in a quantitative fashion. Results The cells were sensitive to both inhibitors but quickly developed resistance that could be inherited, suggesting a genetic origin. Rotation protocols were not useful to postpone the emergence of resistance, which implies that such protocols, or changing from pexidartinib to quizartinib (or vice-versa) should not be used in patients. The computational modelling led to similar conclusions and suggested that F691L is the most common mutation to occur with quizartinib, and also when both drugs are used in rotation. Conclusions AML patients are not likely to benefit from a quizartinib/pexidartinib rotation protocol. A combination of tyrosine kinase inhibitors (with different molecular targets) might be more useful in the future. Development of specific FLT3 inhibitors that are less sensitive to resistance mutations might also lead to a better outcome.

BackgroundAcute myeloid leukaemia (AML) remains difficult to treat despite the development of novel formulations and targeted therapies. Activating mutations in the FLT3 gene are common among patients and make the tumour susceptible to FLT3 inhibitors, but resistance to such inhibitors develops quickly.MethodsWe examined combination therapies aimed at FLT3(+)-AML, and studied the development of resistance using a newly developed protocol. Combinations of FLT3, CDK4/6 and PI3K inhibitors were tested for synergism.ResultsWe show that AML cells express CDK4 and that the CDK4/6 inhibitors palbociclib and abemaciclib inhibit cellular growth. PI3K inhibitors were also effective in inhibiting the growth of AML cell lines that express FLT3-ITD. Whereas resistance to quizartinib develops quickly, the combinations overcome such resistance.ConclusionsThis study suggests that a multi-targeted intervention involving a CDK4/6 inhibitor with a FLT3 inhibitor or a pan-PI3K inhibitor might be a valuable therapeutic strategy for AML to overcome drug resistance. Moreover, many patients cannot tolerate high doses of the drugs that were studied (quizartinib, palbociclib and PI3K inhibitors) for longer periods, and it is therefore of high significance that the drugs act synergistically and lower doses can be used.