2809 Craniospinal Radiation Delivered by Helical Tomotherapy: A Dosimetric Study

M. B. Tomblyn, S. K. Hui, K. E. Dusenbery

University of Minnesota, Minneapolis, MN

Purpose/Objective(s): Craniospinal irradiation (CSI) may be used in both pediatric and adult populations for malignancies witha propensity for CSF spread or known CSF involvement. Unfortunately, the dosimetry of the traditional technique is variablein junctioned regions, and this is exacerbated by the patient’s respiratory cycle. Scheduled junction shifts have been employedin an attempt to avoid overdosing or underdosing in the area of the field matches. Furthermore, the depth of the spinal canalvaries along the length of the posterior field, causing dosimetric variation along the spinal canal. Blocks are constructed for thecranial fields to allow for adequate dose to the cribiform plate while protecting the orbits; however, concern remains for theunderdosing of a critical region of the CNS due to setup error. Small children who may require daily anesthesia present anadditional challenge to treatment in the prone position. Supine positioning addresses the latter concern; however, it suffers fromthe other aforementioned limitations. Ultimately, traditional CSI has inherent dosimetric variations and may be prone to patientset up error. Helical tomotherapy (HT) is a novel image-guided treatment that has been employed to allow for precise targetsculpting while limiting dose to critical normal tissues. It would allow for treatment in the supine position, and structures whoseinclusion in the radiation field is critical to success could be visualized each day prior to treatment. Here, we perform adosimetric comparison of traditional LINAC-based CSI to that delivered by HT.

Materials/Methods: A 25-year-old male with a T2 medulloblastoma, status post gross total resection, was treated to the entireneuraxis using a traditional prone CSI technique to 23.4 Gy, followed by a 30.6 Gy, 3-field conformal boost to the posteriorfossa. Cerrobend blocks were used for the cranial opposed lateral fields, and 1 cm junction shifts were performed after each 10Gy to the spine. Later, the archived CT images were used to generate a 23.4 Gy craniospinal plan using HT, with a 1 cm marginaround the neuraxis to generate a PTV. Inverse planning techniques were employed to constrain dose delivered to normalstructures such as the heart, lungs, kidneys, liver and bowel. Dosimetry to the PTV and critical normal tissues was performedfor both the traditional and HT plans. Traditional CSI fields contained a cold spot representing less than 50% of prescribed doseat the lower spinal junction, where no gaps were necessary using HT. Additionally, HT resulted in a more highly conformaldose delivery to the PTV. Dose inhomogeneity (90% - 10% volume) to the PTV for HT was 3%, compared to 13% for thetraditional CSI setup.

Conclusions: CSI with HT results in superior dosimetry to the PTV, with a higher minimum dose delivered more conformally,while maximum doses to critical normal tissues are lower compared to traditional LINAC-based CSI. HT requires no junctionswith gaps to be calculated and shifts to be made, and it is not subject to critical underdosing as seen in the traditional CSI plan.The patient may be treated in the supine position, making daily anesthesia safer and decreasing the respiratory excursion of thespinal canal. Finally, as an image-guided radiation delivery method, HT allows for setup confirmation to be obtained each dayprior to treatment.

Author Disclosure: M.B. Tomblyn, None; S.K. Hui, None; K.E. Dusenbery, None.

2810 Holographic Image Guided Radiation Therapy (HIGRT) Treatment Planning: a Multi-Institutional Study

J. C. H. Chu1, X. Gong1, M. Kirk1, A. Khan1, M. Rivard2, C. Melhus2, M. Buscher2, G. Cardarelli3, A. Hurley3,J. Hepel3, et al.1Rush University Medical Center, Chicago, IL, 2Tufts-New England Medical Center, Boston, MA, 3Rhode Island Hospital,Providence, RI

Purpose/Objective(s): To determine the feasibility and potential of a novel 3D display device for radiation treatment planning.

Materials/Methods: The Perspecta device displays holographic images by projecting 5,000 frames per second on a spinningscreen. An interface for displaying CT, regions of interest (ROIs), and beams from a modern 3D planning system was developedfor this study. We investigated if this device could improve planning efficiency and/or plan quality. Image and ROI data from12 previously treated brain cancer patents was replanned at 2 institutions. One lung and one breast patient were replanned ata 3rd institution.

Treatment plans were developed using the 3D planning systems with or without Perspecta. Each patient was planned by twodifferent planners to reduce bias. Time required to produce a plan was recorded.

Plan quality was reviewed by 4 physicians. All plans were viewable on both Perspecta and conventional planning systemsat the reviewers’ choice. Perspecta dose cloud display, DVH, and equivalent uniform dose (EUD) were available to assist thereview if desired. Physicians assigned plans to 4 categories (from unacceptable to excellent) based on target volume coverageand critical tissue dose. Reviewing physicians were blinded to the planning device used.

Results: All plans but one were considered acceptable. Brain plans were considered equivalent in 4 patients (33%). Perspectaguided plans were better in 6 patients (50%), because of reductions of up to 24% of brain stem and 28% of optic chiasm volumesreceiving 56Gy. The Perspecta plan was worse in two patients (17%), because of higher EUD to optic nerve and brain stemwith no improvement in target coverage. The Perspecta plans for lung and breast were better than rival plans.

There was no difference in the average time spent to produce a plan, whether or not Perspecta was used, despite the fact thatPerspecta planning tools were not yet fully developed and planners were unfamiliar with their operation. Perspecta displayprovided instant recognition of 3D relationships of dose distribution and anatomy. All reviewers felt that the device allowedbetter appreciation of these 3D relationships than images from a flat screen display. The location and size of over- orunder-dosed regions were also easier to identify on Perspecta.

Conclusions: We have successfully completed a preliminary study to evaluate the usefulness of a holographic display devicein radiation treatment planning. The display provides complex, 3D information in a more efficient and natural way. These datademonstrate that Perspecta has potential to achieve better plan quality in radiation treatment planning even though it does notyet improve planning efficiency. These preliminary results warrant an expanded study to include a larger number of patientsfor a variety of disease sites.

S664 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 3, Supplement, 2006