| Chemotherapy and radiation are common treatments for cancer patients. Their mechanism of action is not targeted to cancerous cells specifically, and non-cancerous cells can suffer the consequences. Too much therapy can induce hematological toxicities. When cell counts dip too low, bleeding and infection can reduce and delay treatment. The use of mathematical models to understand these toxic events is a starting point for avoiding high-grade toxicity. If patients can be flagged as sensitive to a given therapy early on in their treatment, adjustments can be implemented in real-time.;A biologically-motivated model of chemotherapy-induced thrombocytopenia is developed. Platinum compounds such as carboplatin, oxaliplatin, and cisplatin can significantly lower platelet counts. These chemotherapeutics kill stem and progenitor cells in the bone marrow, which ultimately leads to a decrease in the circulating platelet count. The model parameters are able to identify which patients will react strongly to the chemotherapy, as characterized by a rapid decline in cell count followed by a low-magnitude rebound. This model was then expanded to include neutropenia, a separate toxicity common to systemic chemotherapeutics. Clinical outcomes based on survival (good vs. poor) and tumor differentiation from healthy tissue (denoted as high or low delta), were used to bin patients. This model simultaneously captures thrombocytopenia and neutropenia in pancreatic cancer patients and predicts which patients will have a better or worse overall survival time.;While chemotherapy toxicity often affects rapidly dividing cells in the bone marrow, radiation can be toxic whether bone marrow is in the beam path or not. Lymphopenia occurs as a result of the cells circulating through blood vessels in the beam path while the patient is being irradiated. An algorithm was developed to track the amount of integrated damage taken by circulating lymphocytes, and as the damage accumulates which cells die. Lymphocyte death and replenishment occurs at biological rates taken from the literature. This model predicts the depth of lymphopenia after each fraction of radiation and can be combined with a chemotherapy-induced lymphosuppression model to predict cell counts after combination therapy. |
