V. Malkov, L. Buchholtz, S. S. Park, and J. Kavanaugh; Department of Radiation Oncology, Mayo Clinic, Rochester, MN
Purpose/Objective(s): The standard radiotherapy process includes simulation CT for contouring and treatment planning. Using diagnostic CT images instead of sim CT can provide longer planning lead times, reduce patient costs and improve quality of life, but introduces dosimetrically relevant uncertainties such as patient setup differences and HU accuracy of the diagnostic images. Adaptive radiotherapy accommodates for inter-fraction patient setup and anatomical differences and can mitigate many of these uncertainties. This work’s objective is to validate an Ethos adaptive multi-fraction simulation-free (SF) workflow for well selected Stereotactic Body Radiation Therapy (SBRT) patients. Materials/
Methods: The SF workflow utilizes institutional diagnostic PET-CT images to generate an Ethos reference adaptive treatment plan. The diagnostic images are initially reviewed by physician and physicist to evaluate the treatment site, discuss patient challenges, and determine on-treatment patient setup (immobilization, arm position, patient orientation). Beam arrangement is selected to mitigate the impact of limited CBCT field-of-view and regions of uncertainty on the diagnostic image. An adaptive fraction using a Hypersight (HS) capable Ethos system is delivered and the associated CBCT, contours, and plan are used to establish the subsequent non-adaptive fractions. HU and dosimetric accuracy of the PET-CT and Ethos CBCT (HS and non-HS) is performed. The timing of the initial experience with a three-patient cohort is presented. Results: The adaptive-driven SF SBRT workflow was successfully implemented following validation and end-to-end testing. The HU accuracy of the PET-CT images and HS Ethos CBCTs using iCBCT Acuros reconstruction were found to be within the range of the departmental sim CTs. Dose metrics of plans computed on the PET-CTs, Ethos CBCTs, and sim CTs were within 1.2% of the sim CT average of target coverage, OAR near max, and V80% metrics. For the initial three-patients, adaptive delivery times ranged between 50.9 and 37.2 which were comparable to times reported in similar single fraction palliative CT simulation free studies. Careful review of the un-editable auto generated external, and the synthetic CT used for dose computation was performed at each fraction. Preparation time for the transition from adaptive to conventional fractions following the first fractions ranged between 1.3 and 4.2 hours. Subsequent non-adaptive image-guided fractions ranged between 7 and 27 minutes. Conclusion: This study demonstrates the feasibility of a multi-fraction SBRT diagnostic CT simulation free process driven by an initial adaptive fraction to accommodate for patient setup. This work identifies technology development opportunities to facilitate transitioning between adaptive and conventional treatments. SF processes can help reduce on-site patience presence prior to treatment and improve the radiotherapy planning and delivery efficiency.
