Deformable Dose Reconstruction to Optimize the Planning and Delivery of Liver Cancer Radiotherapy

The precise delivery of radiation to liver cancer patients results in improved control with higher tumor doses and minimized normal tissues doses. A margin of normal tissue around the tumor requires irradiation however to account for treatment delivery uncertainties. Daily image-guidance allows targeting of the liver, a surrogate for the tumor, to reduce geometric errors. However poor direct tumor visualization, anatomical deformation and breathing motion introduce uncertainties between the planned dose, calculated on a single pre-treatment computed tomography image, and the dose that is delivered. A novel deformable image registration algorithm based on tissue biomechanics was applied to previous liver cancer patients to track targets and surrounding organs during radiotherapy. Modeling these daily anatomic variations permitted dose accumulation, thereby improving calculations of the delivered doses. The accuracy of the algorithm to track dose was validated using imaging from a deformable, 3-dimensional dosimeter able to optically track absorbed dose. Reconstructing the delivered dose revealed that 70% of patients had substantial deviations from the initial planned dose. An alternative image-guidance technique using respiratory-correlated imaging was simulated, which reduced both the residual tumor targeting errors and the magnitude of the delivered dose deviations. A planning and delivery strategy for liver radiotherapy was then developed that minimizes the impact of breathing motion, and applied a margin to account for the impact of liver deformation during treatment. This margin is 38% smaller on average than the margin used clinically, and permitted an average dose-escalation to liver tumors of 9% for the same risk of toxicity. Simulating the delivered dose with deformable dose reconstruction demonstrated the plans with smaller margins were robust as 90% of patients' tumors received the intended dose. This strategy can be readily implemented with widely available technologies and thus can potentially improve local control for liver cancer patients receiving radiotherapy.Ph.D