RBE of kV CBCT radiation determined by Monte Carlo DNA damage simulations.

Due to the higher LET of kilovoltage (kV) radiation, there is potential for an increase in relative biological effectiveness (RBE) of absorbed doses of radiation from kV cone beam computed tomography (CBCT) sources in reference to megavoltage or Co-60 doses. This work develops a method for accurately coupling a Monte Carlo (MC) radiation transport code (PENELOPE) with the damage simulation (MCDS) to predict relative numbers of DNA double strand breaks (DSBs). The MCDS accounts for slowing down of electrons and delta ray production within the cell nucleus; however, determining the spectrum of electrons incident on the cell nucleus from photons interacting in a larger region of tissue is not trivial. PENELOPE simulations were conducted with a novel tally algorithm invoked where electrons incident on a detection material were tracked and both the incident energy and the final deposited dose were recorded. The DSB yield predicted by a set of MCDS runs of monoenergetic electrons was then looked up in a table and weighted by the specific energy of the incident electron. Our results indicate that the RBE for DSB induction is 1.1 for diagnostic x-rays with energies from 80 to 125 kVp. We found no significant change in RBE with depth or filtration. The predicted absolute DSB yields are about three times lower for cells irradiated under anoxic conditions than the yield in cells irradiated under normoxic (5%) or fully aerobic (100%) conditions. However, oxygen concentration has a negligible (± 0.02) effect on the RBE of kV CBCT x-rays.

Sci-Thur AM: Planning - 09: Assessing dynamic IMRT field modulation in prostate plans.

In a previous study, the variogram fractal dimension (FD) method was found to be very accurate at identifying planned head and neck IMRT fields that are overly-modulated. In the current study, the authors used MATLAB® to develop FracMod, a graphical user interface (GUI) and variogram FD analysis tool to assess modulation complexity of dynamic IMRT fields designed for treatments of the prostate alone and prostate plus pelvic nodes. A set of 5 prostate plans (25 fields) and 5 prostate plus pelvic node plans (35 fields) were used to choose FD cut-points that ensure no false positives (100% specificity) in distinguishing between moderate field modulation (typical modulation used clinically at the authors' institution) and high modulation. Field modulation was controlled by adjusting fluence smoothing parameters in the Eclipse™ treatment planning system. The area under the curve (AUC) from receiver operating characteristic (ROC) analysis was used to quantitatively compare the ability of FD and the number of monitor units (MUs) for distinguishing between the moderate and high modulation fields. The variogram FD method gave AUCs of 0.96 (almost perfect classification) and 1.00 (perfect classification) for the prostate alone and the prostate plus pelvic node fields, respectively. The variogram FD method is an accurate metric; performing better than the number of MUs at identifying high modulation IMRT fields planned for the treatment of prostatic carcinoma. Hence, FracMod will enable Radiotherapy Physicists to easily and accurately quantify the degree of modulation of IMRT fields and adjust overly-modulated fields at the treatment planning stage.

Sci-Fri PM: Delivery - 08: Characterizing the spatially varying fluence and spectra of a kV imaging source for dose calculations.

Kilovoltage (kV) daily image-guided radiotherapy (IGRT) procedures accumulate radiation dose within the patient that is currently not routinely incorporated in the treatment plan. As part of the process of developing a patient-specific kV dose computation tool, the kV x-ray source must be characterized. We propose a simple, clinically feasible experimental characterization method using in-air dose measurements along the transverse axis. We determine half-value layer (HVL) along the transverse axis, from which we derive the HVL-specific mass-absorption coefficient, which is used to determine beam fluence. These values are interpolated over the entire field. The spectrum at each interpolation point in the field is found from HVL and accelerating potential (kVp) using third-party software Spektr. We use this method to characterize the spatially varying fluence and spectra of a Varian® On-Board Imaging® source for energies 80, 100 and 125 kVp. This characterization is used to compute dose within a heterogeneous phantom, using our previously validated in-house dose computation software, which we compare with relative dose measurements. We show that for a 10×10 cm2 field size using no added filtration, the agreement for all three energies is within 2% for the central depth-dose profile and within 2.6% for the transverse profiles. This clinically feasible experimental characterization method for kV imaging sources represents a crucial step in the development of a patient-specific dose computation tool.

Sci-Sat AM: Brachy - 06: Monte carlo DNA damage simulations of kV cbct radiation.

When performed daily, cone beam CT (CBCT) images can accumulate radiation dose to non-negligible levels. Because kV x-rays have a larger relative biological effectiveness (RBE) than its MV x-rays, the accumulated absorbed dose needs to be multiplied by an appropriate RBE to better evaluate the impact of CBCT dose in a treatment planning context. We investigated this question using PENLEOPE simulations to look in detail at the electron energy spectra produced by kV x-rays and Co-60 γ-rays in biologically motivated geometries. The electron spectra were input into the published Monte Carlo Damage Simulation (MCDS) and used to estimate the average number of double strand breaks (DSBs) per Gy per cell. Our results suggest an approximately 10% increase in the RBE for DSB induction. For the majority of treatment planning scenarios where imaging dose is only a small fraction of the total delivered dose to target volumes and organs at risk, the increase in RBE is not critical to be factored in, however for it may play a significant role in predicting the induction of secondary cancers.

Flow patterns at the stenosed carotid bifurcation: effect of concentric versus eccentric stenosis.

Carotid stenosis severity is a commonly used indicator for assessing risk of stroke. However, the majority of individuals with severe carotid artery disease never suffer a stroke, and strokes can occur even with only mild or moderate stenosis. This suggests local factors (other than stenosis severity) at or near the carotid artery bifurcation may be important in determining stroke risk. In this paper we investigate the effect of stenosis geometry on flow patterns in the stenosed carotid bifurcation, using concentrically and eccentrically stenosed anthropomorphic carotid bifurcation models having identical stenosis severity. Computational simulations and experimental flow visualizations both demonstrate marked differences in flow patterns of concentric and eccentric stenosis models for moderately and severely stenosed cases, respectively. In particular, we identify post-stenotic recirculation zone size and location, and spatial extent of elevated wall shear stress as key factors differing between the two geometries. As these are also rotid plaque more vulnerable to cerebral embolus prokey biophysical factors promoting thrombogenesis, we propose that the stenosed carotid bifurcation geometry--or the induced flow patterns themselves--may provide more specific indicators for those plaques that are vulnerable to enhanced thromboembolic potential, and hence, increased risk of ischemic stroke.