Outcome prediction and personalised therapy for leukaemia
In the field of cell therapy, culture models used to investigate tumour cell behaviour and responsiveness to therapies, are almost exclusively two-dimensional. They fail to capture the three dimensions of the bone marrow architecture and therefore generate only low-quality cells. BioBlood fills this gap with a new generation of 3D culture platforms. “Our platforms offer a dynamic, 3D perfusion bone marrow biomimicry in a bioreactor. It produces multiple cell types simultaneously in an environment that is both serum-free and cytokine-free,” explains Prof. Athanasios Mantalaris, PhD at Imperial College London and coordinator of BioBlood. The stromal cells produced by BioBlood’s platforms are based on input cord blood cells like those observed in bone marrow. One cord blood unit can be sustained in continuous dynamic culture for at least 8 weeks. This makes BioBlood’s personalied ex vivo platform a viable solution for the production of blood cell components. In the future, it could be used for transfusion purposes or to discover personalised drug targets. Towards personalised chemo- and immunotherapy Personalised chemotherapy, on the other hand, is still faltering. Current doses for standard treatments are administered based on height, weight and performance status of the patient. But so far, they have failed to consider leukemia cell kinetics, or even how resistant mutations and microenvironmental factors could affect such cell kinetics. BioBlood overcomes these limitations with the first-ever, in silico model for precision therapeutics and optimisation of treatment schedule and dose. “We started this project by trying to improve the efficacy and safety of treatments. We have used in silico modeling of parameters obtained during routine diagnostic testing on patients with Acute Myeloid Leukaemia (AML). Eventually, we could combine patient-specific and leukaemia-specific parameters with the pharmacodynamics and pharmacokinetics of standard chemotherapy drugs. We were also able to combine these elements with specific actions of chemotherapy drugs on the cell cycle, as well as account for the heterogeneity of different populations of normal cells as well as leukemic blasts,” says Prof Mantalaris. From data acquired during patient diagnosis, BioBlood’s mathematical in silico model can determine the response to therapy (complete remission, partial remission, relapsed disease and resistant disease). It can capture neutrophil dynamics during all cycles of chemotherapy, as well as help optimise treatment schedules and doses for improved effectiveness of the treatment and reduced toxicity. “These outcomes were determined through retrospective datasets obtained from patients treated for AML. We are now in the planning phases of a prospective clinical trial to assess whether these outcomes can be dynamically predicted. If they can, this in silico model could lead to a step-change in how AML will be treated in future. It would allow for dynamic scheduling to increase chemotherapy efficacy and reduce toxicity,” says Prof Mantalaris. The in silico precision therapy platform can incorporate both standard chemotherapy and novel immunotherapies into its mathematical model. Whilst Brexit has jeopardised chances for an EU-funded follow-up project, the consortium has created a spin-out company called πiChemo and intends to focus on the US market.
Keywords
BioBlood, leukaemia, personalised medicine, cell therapy, chemotherapy, AML, 3D culture, in silico model, toxicity
