ESTRO 2021 Abstract Book

S1299

ESTRO 2021

Figure 2. Histogram of the displacement of surgical clips relative to breast/chest and lymph node target regions as measured on 1682 daily CBCT scans. Corresponding systematic (Σ) and random components (σ ) is shown in each subfigure. Conclusion Displacement of the surgical clips in the boost region is a major uncertainty in simultaneous integrated boost for breast cancer radiotherapy. However, with the use of daily CBCT performing a match on breast/chest and lymph node regions, the additional uncertainty from clips displacement alone can be accounted for by adding a PTV margin to the boost region of 5mm. PO-1574 Evaluation of automated plan quality for cervical cancer using the Ethos TPS M. Bolt 1 , C. Shelley 2 , R. Hollingdale 1 , S. Chadwick 1 , A. Barnard 1 , A. Leverton 1 , A. Stewart 2 , E. Adams 1 , C. South 1 1 Royal Surrey, Radiotherapy Physics, Guildford, United Kingdom; 2 Royal Surrey, Oncology, Guildford, United Kingdom Purpose or Objective Online adaptive radiotherapy (oART) with daily re-planning requires accurate, robust, high-quality treatment plans to be created in a matter of minutes. Here we evaluate automated plan (AP) quality of the Varian Ethos TPS against manual plans (MP) for 10 cervical cancer patients. Materials and Methods A full-bladder CT scan was used to plan 12-field IMRT to a median PTV dose of 50.4Gy/28#. The Ethos AP template was iteratively optimised for 6 cervical cancer cases, then tested on 10 separate cases. The template included all clinical objectives plus additional goals to aid the optimiser in reducing OAR doses and increasing target coverage. AP plans were generated with no additional user input. MPs were created in Eclipse by experienced radiotherapy planners. All plans were calculated using the Acuros algorithm. Local clinical DVH constraints were used to assess acceptability. Both APs and MPs were physician reviewed and rated pass/minor deviation/fail. Timings for plan setup and optimisation were recorded for MP generation. Results Plan setup for MP took a mean of 15 minutes (range 9-20). The number of optimisations ranged from 3 to 15 with total mean planning time of 133 minutes (range 38 – 360). Less than 15 minutes of user interaction was required to generate APs. Fig.1 shows the range of OAR mean doses for each TPS. Rectal and bone marrow doses reduced (p<0.05) with AP from 45.3Gy (40.0-49.3Gy) to 43.7Gy (37.5-48.8Gy) and from 30.1Gy (27.0–32.3Gy) to 28.4Gy (23.4–30.2Gy) respectively.