Single high dose irradiation: applicability of cell survival curve modeling and in vivo evaluation of tumor biologic response
Date
2015
Authors
Swancutt, Katy Lynn, author
LaRue, Susan M., advisor
Custis, James T., committee member
Liber, Howard L., committee member
Page, Rodney L., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
The observed clinical success of stereotactic radiation therapy (SRT), in which radiation therapy treatment is delivered in one to five, high-dose fraction of ionizing radiation, has generated interest in the biologic mechanisms by which SRT achieves tumor control. Since the use of linear quadratic formalism to predict tumor cell kill and clinical tumor control has not been corroborated by clinical trial results, it has been proposed that alternative mechanistic responses occur in response to high dose irradiation in addition to mitotic cell death of tumor clonogens in direct response to radiation-induced DNA damage. One suggested mechanism based on observations of tumor endothelial cell apoptosis following doses above 8 to 10 Gy proposes that tumor vascular damage may increase cell killing of associated tumor parenchyma, thus explaining the clinical success of SRT. The work described herein sought to determine whether tumor cell killing behaved predictably at high doses based on established survival curve models using in vitro techniques and whether parameters to tumor vascular response in vivo were impacted differently at high doses versus low doses. In addition, the role of vascular dysfunction as a potentially impactful process in the tumor response to high dose, single fraction irradiation was investigated by measuring microenvironmental changes in the time period between irradiation and the expected peak of endothelial cell apoptotic events. For both in vitro and in vivo experiments, canine tumors were chosen as the optimal model for human cancers due to the similarities between canine and human tumor size, cellular kinetics, and initiation from similar environmental exposures (among numerous other reasons). Using the standard clonogenic assay, cell survival curves were constructed for eight different canine cancer cell lines and one normal canine endothelial cell line. Careful conduction of experiments allowed for collection of data at doses above 10 Gy (in many cases up to 14 or 15 Gy). Modeling of these data was conducted using the linear quadratic formula, the well-established single-hit multitarget model, the new Kavanagh-Newman universal survival curve, and a hypothetical model based on the linear quadratic model with an added component proportional to dose cubed. It was determined that no single model provided the best fit for all cell lines and the linear quadratic model failed to describe data in the high dose region in a majority of cell lines. Rank order analysis of cell lines showed consistency between measured surviving fractions in the low dose region (SF2 versus SF8), but extrapolation of measured high-dose data to calculate surviving fraction values at 24 Gy (SF24) produced rank orders nearly opposite of those at low doses. The reversal of ranks indicated crossing of survival curves at some point beyond the limits of experimental measurement, suggesting that perhaps some unique mechanisms (in isolated tumor cells, in the absence of stromal components) occur at high doses that have not been observed. Extrapolated SF24 values, representing cell survival following a single fraction of 24 Gy, were compared to surviving fraction calculations for a dose fractionation schedule of three fractions of 8 Gy to total 24 Gy by calculating the cubed value of each cell line’s SF8. The in vitro studies indicated that a single fraction of 24 Gy resulted in several orders of magnitude more cell killing than three fractions of 8 Gy, leading to the prediction that a single 24 Gy treatment would be more clinically efficacious than a treatment delivering 24 Gy divided among three equal fractions. Such in vitro studies may be useful in guiding the design of dose fractionation protocols for SRT. To investigate the role of tumor vasculature in tumor response to single fraction, high-dose irradiation, spontaneously occurring canine soft tissue sarcomas were treated with either 2, 8, or 24 Gy in a randomized clinical trial. Tumor microenvironmental factors related to vascular function were monitored before and after irradiation. These factors included tumor oxygenation, interstitial fluid pressure, and perfusion. Samples were taken for quantification of endothelial cell apoptosis, plasma nitrate and nitrite as long-lived indicators of nitric oxide, serum ASMase and oxidative stress. The timeline of events in which high dose, single fraction irradiation induced changes in the tumor microenvironment was of particular interest, and measurements were taken for three hours immediately after irradiation and again at twenty four and forty eight hours after irradiation. Interim analysis of the clinical trial following enrollment of nine dogs suggests that oxygenation following a single fraction of 2 Gy behaved consistently with expectations at twenty four and forty eight hours. Tumors treated with 24 Gy showed a trend of increasing oxygenation within one to two hours of irradiation, which decreased again by three hours. It was concluded that more dogs are needed to clarify any trends observed in this study, given the high degree of heterogeneity of tumor microenvironmental factors within a tumor and between individuals.