Challenging the Rigor of Alpha/Beta and Relative Biological Effectiveness in Tumor Contorl [sic] Probability Determinations via In Silico Modeling That Incorporates Individualized Biological and Physiological Data

BACKGROUND: Current methods for predicting the response of tumors to radiation therapy focus largely on linear quadratic (LQ) model survival parameters generated from conventional radiotherapy regimens, of ~2 Gy/fraction. The accuracy of the LQ model has been questioned for high dose per fraction radiation therapy and for carbon ion radiotherapy. Furthermore, the determination of relative biological effectiveness (RBE) values for charged particles is based upon the survival response of a very limited number of mammalian cell lines. Such potential discrepancies in the determination, or even the utility of, the LQ parameters and RBE values may negatively impact tumor control. OBJECTIVES: This dissertation expands upon pre-existing models that calculate tumor control probability (TCP) by adding additional biologic and physiologic metrics as well as incorporating alternative radiation survival model parameters, such as the repairable-conditionally repairable (RCR) model. The primary objective is to determine the extent to which these different approaches to determining TCP may impact predictions of tumor cure dose for specific dose fractionation schedules for photons as well as determine the extent to which TCP varies when personalized RBE values for carbon ion irradiations are used compared to TCP determinations when generalized RBE values are employed. METHODS: An existing in silico model was expanded by incorporating biological information such as intrinsic radiosensitivity, hypoxia, tumor repopulation after irradiation, number of cells per tumor, and physiological information such as whether a tumor undergoes reoxygenation between fractions. TCP curves represent model output and both the LQ and RCR parameters for a given fractionation schedule. A comparison of the efficacy of hypofractionated radiotherapy and carbon ion irradiation was made through relative clinical effectiveness (RCE) which incorporates the dose to reach TCP of 50% (TCP50) as an endpoint between reference and test treatment schedules. RESULTS: The LQ model suggest that it takes less dose to reach TCP50 compared to the RCR model, indicative of the concerns that the LQ model overpredicts cell killing at higher doses per fraction. Tumor response modeled with conventional fractionation (2 Gy/fx) schemes was unsuccessful in that TCP was determined to be 0, however, where treatment schedules employing 8 Gy/fx of photon radiation or carbon ion irradiation predicted TCP values were low but non-zero and were correlated with the intrinsic radiosensitivity of the cell line. RCE was higher for tumors with intrinsic radioresistance to photons when treated with hypofractionated photon or carbon ion irradiation demonstrating that tumors resistant to standard fractionated radiotherapy could benefit more from alternative modalities. CONCLUSIONS: The in silico modeling performed here suggests that the selection of survival model to quantify radioresponse made a profound difference in the in silico TCP calculations, as did personalized biological and physiologic information. Furthermore, using personalized RBE values in TCP determination significantly influenced the determination of RCE values. The approach implemented here could serve as a method of triage for tumors with a radioresistant phenotype. With time it may be possible to identify patients whose tumors hide a radioresistant phenotype and direct those individuals to more appropriate therapies.

Author dc:creator

• Polsdofer, Elizabeth Marie

Contributors dc:contributor

• Wang, Jing
• Story, Michael
• de Gracia Lux, Caroline
• Davis, Anthony John
• Jia, Xun

• Cell Survival
• Dose Fractionation, Radiation
• Neoplasms
• Radiobiology
• Relative Biological Effectiveness
• Cell Death

Language dc:language

en

Identifier

1596185284

OAI identifier oai:identifier

oai:utswmed-ir.tdl.org:2152.5/10825