A mathematical model of cell cycle heterogeneity for personalizing leukemia chemotherapy

• Main Author: Fuentes Gari, Maria
• Other Authors: Mantalaris, Athanasios; Pistikopoulos, Efstratios N.
• Published: Imperial College London 2015
• Subjects: 660
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.666506

Acute myeloid leukemia is a type of blood cancer characterized by an excessive build-up of immature blood cells in the bone marrow and blood streams. As a result, healthy stem cells become space- and resource-limited, and do not produce enough functional cells for the body to operate normally. Treatment is required immediately, consisting of intensive chemotherapy. Chemotherapy dosage and schedule are derived from established protocols which do not account for patient-specific and disease-specific heterogeneity. Over- or under- dosage are thus common; a more rational and personalized approach to chemotherapy treatments is required. Specifically, incorporating the effect of chemotherapy in a cell cycle phase-specific manner would be highly beneficial. In this work, we developed a population balance model (PBM) of the cell cycle based on the underlying biology that captures the progress of cells within and between phases. It was validated with three leukemia cell lines separately for the duration of one cell cycle, and in variable mixtures where forward and backward kinetics as well as clonal identification were successfully performed. The model was compared against two other cell cycle models: an existing ODE model and a newly developed DDE model featuring phase durations as delays. The PBM outperformed the other two in recapitulating biological features, and displayed a higher sensitivity to treatment when coupled to an existing pharmacokinetic/pharmacodynamic model of chemotherapy treatment. The PBM was further used in the prediction of clonal evolution during chemotherapy, highlighting the important heterogeneity in treatment response between clones but also the competitive features among them that could be critical in the success of the treatment. Finally, the first steps towards implementing this technology at clinical level were taken by defining converted, measurable data sets. A prototype application, "ChemoApp", was developed at the user interface level for the introduction of this research into clinical practice.