'CellTrajectory' for cellular automata modelling of leukaemic stem cell dynamics in acute myeloid leukaemia: insights into predictive outcomes and targeted therapies

Acute myeloid leukaemia (AML) is a haematologic malignancy with high
relapse rates in both adults and children. Leukaemic stem cells (LSCs) are
central to leukaemopoiesis, treatment response, and relapse and frequently
associated with measurable residual disease (MRD). However, the dynamics
of LSCs within the AML microenvironment are not fully understood. This
study utilized three-dimensional cellular automata (CA) modelling to
simulate LSC behaviour and treatment response under induction
chemotherapy. Our study revealed: (1) a correlation between LSC
persistence post-induction chemotherapy and risk of AML relapse; (2) MRD
negativity based on LSC count may not reliably predict outcomes,
supporting clinical evidence that patients with MRD-negative status can
still be at risk of relapse; (3) prolonged persistence of LSCs
post-chemotherapy without disruption of normal haematopoiesis, aligning
with clinical observations of dormant AML clones; (4) early LSC dynamics
post-induction chemotherapy, characterized by stochastic behaviours and
movement velocities, are insufficient predictors of long-term prognosis;
and (5) a distinct spatiotemporal organization of LSCs in later phases
post-induction chemotherapy is correlated with long-term outcomes. Our
modelling results provide a theoretical and clinical framework for AML
research, and future clinical data validation could refine the utility of
CA modelling for oncological studies.