School of Mathematics

Giulio Caravagna

Timing the origin of immune-escape from leukaemia after bone marrow transplant

In treating Acute Myeloid Leukaemia (AML), a bone marrow transplant introduces healthy donor stem cells to replace the leukemia-afflicted marrow, aiming to regenerate a healthy blood and immune system. Transplant works by triggering an immune response where donor immune cells attack residual leukaemia cells, beneficially contributing to AML treatment. However, AML cells can evade this immune attack through various mechanisms allowing them to hide from the immune system. In some type of transplants, immune evasion can be achieved by a complex structural rearrangement of the chromosome encoding for the locus used by the immune system to recognise AML cells. Like in all treatment-resistant scenarios, one of the most important question is understanding whether the treatment-resistant population was born before, or after treatment. In this talk I will approach this question thanks to a combination of longitudinal high-resolution whole-genome sequencing data, combined with mathematical models from Poisson processes, population genetics and Bayesian inference. By modelling the complex interplay between genomic mutations, structural alterations and drug-exposure, we were able to time the origin of the treatment-resistant clone in 10 distinct patients. For the first time, these results shed light on a complex process of immune-evasion using advanced computational modelling, focusing on one of the most aggressive type of leukaemia and offering a paradigmatic approach to be translated to other types of cancer.