On-Board Cone Beam Computed Tomography as Radiation Therapy Simulation for Emergent Cases

Abstract

To investigate dosimetric clinical appropriateness of on-board kV cone beam computed tomography patient simulation (CBCT-Sim) for palliative treatment plans at a remote satellite clinic having no conventional computed tomographic simulation (CT-Sim) when travel isn't a viable option. Our current approach is a separation calculation using two opposing fields on a water phantom. A CBCT of the patient would improve the plan by utilizing the patient's actual geometry and we expect to see dose points in the CBCT-Sim plan image set less than 3% different than a conventional CT, and the separation calculation to be within 5% of the conventional CT-Sim plan. Patients treated for the four most likely emergent or immobile disease sites (brain, thoracic spine, lumbar spine, and pelvis) with an original CT-Sim and at least one CBCT during course of treatment were investigated retrospectively. Three brain patients, three T-spine, three L-spine, and two pelvis patients were used. The patients’ images were copied into a new, anonymous patient. A new CT-density curve was created in the treatment planning software for the CBCT based on CT-density phantom data. A standard plan for each given site was planned on the original CT image set, then copied on both a heterogenous and homogenous CBCT image set. The homogenous CBCT image was density overridden to water. A rectangular water phantom was also created for a separation calculation plan, which is our standard practice for these cases. For the brain cases, the CT and heterogenous CBCT plans were both contoured with critical structures. For all cases, seven to eleven points of interest in all directions (isocenter, and fractional separation distances in 3-dimentional space) were used to compare both CBCT-Sim plans and the separation calculation to the standard CT plan. For all cases, the average dose difference per case compared to the standard CT-Sim plan was less than 3% for either CBCT-Sim plan and less than 4% for the separation calculation plan; however, the separation calculations showed large deviations, up to 7.2%, while the CBCTs showed a maximum of 4.8%. Between all cases, the average dose difference was lowest compared to CT-Sim plans for the heterogenous CBCT plan with 1.2±0.7%, while the homogenous CBCT plan difference was 1.4±0.7% and the separation calculation was 1.7±1.1%. The most dosimetric benefit for the heterogenous CBCT-Sim plans was seen for the brain plans, followed by the T-spine plans, with negligible benefit for pelvis and L-spine plans. Dosimetric accuracy is improved with a CBCT-Sim compared to our standard separation-based plan, due to the use of actual patient geometry and density.