2910 - Radiobiology Population Modeling of Intestinal Stem Cells and Intestinal Epithelial Cells to Predict Radiation Induced Toxicity

J. Bebawy¹, Y. Menon¹, H. M. McGee², and W. T. Watkins²; ¹University of California Riverside, Riverside, CA, ²Department of Radiation Oncology, City of Hope National Medical Center, Duarte, CA

Purpose/Objective(s): Classical radiation biology including the linear quadratic model do not consider tissue regeneration and the co-dependence of stem cell and differentiated cell populations. Utilizing a set of coupled nonlinear differential equations, we introduce a new radiobiological model which accounts for the co-dependent relationship between intestinal stem cells (ISCs) and intestinal epithelial cells (IECs) and accounts for their unique radiation sensitivity. We validate the model for prediction of radiation-induced toxicity. Materials/

Methods: Model parameterization was performed for optimization including a dose dependent ISC growth parameter which collapses ISC generation at high doses during acute and fractionated radiation schedules. A transition function to represent ISC generation of IECs is a linear function of the IEC population, and an apoptotic ISC term assumes the average ISC lifetime is 60 days. The transition function from ISC to IEC generates a ratio of 250 IECs to 5 ISC per crypt in the steady state, and an apoptotic term assumes IEC survival is 5 days. Dose dependence is included in the ISC growth factor in addition to classical exponential cell loss with ISC-a=0.35/Gy and IEC-a=0.2/Gy. A stable steady state of ISC and IEC populations was perturbed to model dose-response and complications including a single radiation exposure of 3 Gy, 5 Gy, and 10 Gy to model the Gastro-Intestinal Syndrome (GIS), a 5-week 25-fraction (fx) radiation schedule of 1.6 Gy/fx, 1.8 Gy/fx, and 2.0 Gy/fx, and an every-other-day 5-fx stereotactic body radiation therapy (SBRT) schedule of 6-7 Gy/fx.
Results: Perturbation of the steady state model with 3 and 5 Gy in a single exposure result in continued steady state survival of ISC and IEC populations, albeit at lower ratios. A single 10 Gy fraction collapses the model within 3 weeks of irradiation, successfully modeling GIS and the potential for lethal toxicity. Five-week fractionated schedules show stable and rapid recovery of 1.6 Gy/fx, a slower recovery of 1.8 Gy/fx after completion of radiation, and a collapse of populations at doses >2 Gy/fx consistent with complications following >50Gy in 25 fx. For SBRT treated in 5-fx, every-other-day including one weekend, a 6Gy/fx dose is recoverable but 7 Gy/fx results in population collapse representing clinical SBRT complications that are observed in the abdomen and pelvis.
Conclusion: A radiobiology model which includes co-dependent stem and differentiated cell populations, their unique radiosensitivity, and dose-dependent perturbations in the growth rate of stem cells is crucial to modeling long-term complications and advancing our ability to predict outcomes from personalized radiotherapy. Future directions will include unique characteristics of the small and large intestine and parameterize the model to improve prediction of the timing and severity of complications.