Cavity theory applications for kilovoltage cellular dosimetry

Relationships between macroscopic (bulk tissue) and microscopic (cellular) dose descriptors are investigated using cavity theory and Monte Carlo (MC) simulations. Small, large, and multiple intermediate cavity theory (SCT, LCT, and ICT, respectively) approaches are considered for 20 to 370 keV incident photons; ICT is a sum of SCT and LCT contributions weighted by parameter d. Considering μm-sized cavities of water in bulk tissue phantoms, different cavity theory approaches are evaluated via comparison of Dw,m/Dm,m (where Dw,m is dose-to-water-in-medium and Dm,m is dose-to-medium-in-medium) with MC results. The best overall agreement is achieved with an ICT approach in which d=(1−e−βL)/(βL), where L is the mean chord length of the cavity and β is given by e−βRCSDA=0.04 (RCSDA is the continuous slowing down approximation range of an electron of energy equal to that of incident photons). Cell nucleus doses, Dnuc, computed with this ICT approach are compared with those from MC simulations involving multicellular soft tissue models considering a representative range of cell/nucleus sizes and elemental compositions. In 91% of cases, ICT and MC predictions agree within 3%; disagreement is at most 8.8%. These results suggest that cavity theory may be useful for linking doses from model-based dose calculation algorithms (MBDCAs) with energy deposition in cellular targets. Finally, based on the suggestion that clusters of water molecules associated with DNA are important radiobiological targets, two approaches for estimating dose-to-water by application of SCT to MC results for Dm,m or Dnuc are compared. Results for these two estimates differ by up to 35%, demonstrating the sensitivity of energy deposition within a small volume of water in nucleus to the geometry and composition of its surroundings. In terms of the debate over the dose specification medium for MBDCAs, these results do not support conversion of Dm,m to Dw,m using SCT.